# PIM Atlas: full text

Source: https://pimatlas.com/ · Last verified 2026-10-17 · Licence: cite with a link to the page URL.

## Vendor: 4ALLPORTAL
URL: https://pimatlas.com/vendors/4allportal/
Product: 4ALLPORTAL DAM and PIM | Category: DAM and PIM suite | HQ: Gütersloh, Germany | Founded: 2001 (vendor states it has served customers since 2001) | Ownership: Private (4ALLPORTAL GmbH); ownership not disclosed
Deployment: SaaS, On-premise | Segments: Mid-market, Enterprise | Standards: None verified
Pricing model: Quote-based and individually configured by modules, integrations, hosting and services; no self-service pricing. A 14-day trial system is offered.

4ALLPORTAL is a German DAM and PIM suite from Gütersloh that lets organisations manage media assets and product data in one modular platform. It runs as SaaS hosted in Germany or the US, on-premise or hybrid, and offers a wide connector catalogue spanning SAP, Microsoft Dynamics 365, Shopware, Shopify and Adobe InDesign. Its strongest fit is mid-sized and larger companies where asset management is as important as product data, such as consumer brands and media-heavy manufacturers.

4ALLPORTAL GmbH is a software company in Gütersloh, North Rhine-Westphalia, led by managing directors Daniel Lücke and Dominic Vieregge. The vendor states it has helped companies since 2001 and develops its software in Germany. Its platform combines digital asset management, product information management and master data management modules, which can be bought separately and extended later.

The PIM covers variants, product families and technical specifications, imports and harmonises data from ERP systems, Excel files and manufacturers, and runs rule-based validation and completeness checks with audit trails. Assets are linked to products, variants and families, so one change updates every connected item. A published connector catalogue lists ERP systems (SAP, Microsoft Dynamics 365, Infor, Sage, Oracle), shop systems (Shopware, Shopify, Magento, WooCommerce), CMS platforms, Adobe Creative Cloud, print tools such as priint:suite and EasyCatalog, and 1WorldSync (atrify). Typical implementation takes two to four quarters.

Published references include bofrost, which runs 21 PIM modules across 11 countries, medizinfuchs with more than 350,000 products, Teekanne, CEWE and Warner Bros. Discovery. In 2026 the vendor released role-based dashboards (Core version 5.11) and a natural language AI search. AI features also include auto-tagging, OCR search, DeepL translation and OpenAI-based product text generation.

Strengths: (+) DAM and PIM on one platform, with assets linked directly to products, variants and families (+) Broad published connector catalogue covering SAP, Microsoft Dynamics 365, Shopware, Shopify, Adobe and 1WorldSync (+) Choice of SaaS hosted in Germany or the US, on-premise or hybrid (+) Shipped AI features across DAM and PIM, including natural language search, auto-tagging, DeepL translation and OpenAI text generation (+) Multi-country references such as bofrost across 11 countries
Considerations: (-) DAM heritage means buyers with complex B2B classification needs should test PIM data modelling depth (-) Support for ETIM, ECLASS, BMEcat or GS1 is not documented on the PIM product page (-) No public pricing; quotes depend on modules, integrations and hosting (-) Implementation typically takes two to four quarters according to the vendor
Best for: Mid-sized and larger brands and manufacturers that need strong digital asset management and product data management from a single German vendor.

## Vendor: Agility PIM (insightsoftware)
URL: https://pimatlas.com/vendors/agility-multichannel/
Product: Agility PIM | Category: Mid-market PIM | HQ: London, United Kingdom | Founded: 2010 (incorporation of Agility Multichannel Limited) | Ownership: Part of insightsoftware, which acquired Magnitude Software in 2021; insightsoftware is private-equity backed
Deployment:  | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based; no public price list.

Agility PIM is a product information management and PXM platform owned by insightsoftware, the private-equity-backed reporting and analytics software group. It originated with UK-based Agility Multichannel and serves manufacturers, distributors and retailers that publish to web, print and marketplace channels. Public product detail is thin, so buyers should expect to verify deployment options, integrations and standards support directly in a scripted demo.

Agility PIM is sold under its own brand at agilitypim.com, while the site's legal pages and support sit with insightsoftware. The product traces back to Agility Multichannel Limited, a London-registered software company incorporated in 2010. insightsoftware acquired Magnitude Software, the group that marketed Agility PIM, in a deal announced on 28 October 2021, and lists Agility PIM within its product portfolio.

The platform centralises product data, enriches it, automates supplier onboarding and normalisation and publishes to websites, mobile, print and electronic catalogues, kiosks and marketplace feeds. Agility PIM highlights AI-driven enrichment and automatic generation of product names and descriptions. It also publishes material on agentic commerce and on preparing Digital Product Passport data from PIM records.

Customer logos on the vendor site include Avon, Teledyne FLIR, LKQ, Ryobi and ASSA ABLOY, and customer stories cover Sparex, ESPO and Axminster Tools. Because Agility PIM is one product within a large analytics and reporting group, buyers should ask about the dedicated PIM roadmap, the size of the PIM engineering team and the implementation partner network in their region before shortlisting.

Strengths: (+) Long-established PIM with UK origins and named manufacturing, distribution and retail customers (+) Print and electronic catalogue publishing alongside web and marketplace channels (+) Supplier onboarding and data normalisation functions aimed at distributors (+) Backing of insightsoftware, a large software group with reporting and analytics products
Considerations: (-) Very limited public documentation of deployment model, integrations and standards support (-) PIM is a small part of insightsoftware's finance and reporting-focused portfolio, so roadmap investment should be checked (-) No published pricing (-) AI features are described in general terms without named functions or providers
Best for: UK and North American manufacturers and distributors with print and web catalogue requirements that are comfortable validating an established PIM through detailed reference calls.

## Vendor: Akeneo
URL: https://pimatlas.com/vendors/akeneo/
Product: Akeneo Product Cloud | Category: Enterprise PIM | HQ: Nantes, France | Founded: 2013 | Ownership: Private, venture and growth-equity backed (Summit Partners led the 2019 Series C and the US$135 million Series D in 2022)
Deployment: SaaS, On-premise | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based SaaS subscription; no public list prices were verified for this profile. The open source Community Edition is free to download and self-host without vendor support.

Akeneo is a French PIM (product information management) vendor whose Akeneo Product Cloud combines PIM, supplier data onboarding, DAM, syndication through Akeneo Activation, PX Insights and, since the June 2026 PricingHUB acquisition, pricing. It suits mid-market and enterprise brands, manufacturers and retailers that want a SaaS PIM with a large app ecosystem. Gartner named Akeneo a Leader in its 2026 Magic Quadrant for Product Information Management Solutions, and its Agentic Ziggy layer coordinates AI agents for enrichment and data quality.

Akeneo was founded in 2013 in Nantes, France, and built its reputation on an open source PIM released under the Open Software Licence 3.0. The commercial offer is now the Akeneo Product Cloud, a SaaS platform that groups Akeneo PIM, Supplier Data Manager, Akeneo DAM, Activation for syndication to more than 500 retailers and marketplaces, Shared Catalogs, Digital Showroom and PX Insights. The company is privately held: Summit Partners led its Series C in 2019 and its US$135 million Series D in 2022.

AI has moved to the centre of the roadmap since the 2023 acquisition of Unifai. The Winter Release of January 2026 added a native MCP server and a Stripe Agentic Commerce Suite partnership. The Summer Release of July 2026 introduced Agentic Ziggy, an orchestration layer for specialised agents with approval controls and visibility into agent actions. In June 2026 Akeneo acquired PricingHUB, which runs as a dedicated business unit and adds pricing management to the platform.

Buyers running self-hosted commercial Akeneo should note that Flexibility version 7 (PaaS and on-premise) reached end of support on 30 September 2026, according to Akeneo's supported versions table, which leaves SaaS as the main commercial route. The Community Edition remains available as open source for self-hosting, without vendor service levels. Gartner named Akeneo a Leader in the 2026 Magic Quadrant for PIM Solutions.

Strengths: (+) Named a Leader in the 2026 Gartner Magic Quadrant for Product Information Management Solutions (report dated 12 October 2026). (+) Broad SaaS suite: PIM, Supplier Data Manager, DAM, Activation syndication to 500+ retailers and marketplaces, Shared Catalogs and PX Insights. (+) App Store with more than 200 listed integrations, including Shopify, Salesforce, SAP Commerce and Bynder. (+) Shipped agentic AI with governance controls: Agentic Ziggy, a native MCP server and AI image editing in Akeneo DAM. (+) Large open source user base and community knowledge from the Community Edition (OSL 3.0).
Considerations: (-) Flexibility (PaaS and on-premise) version 7 support ended on 30 September 2026; self-hosted commercial customers need a migration plan to SaaS. (-) Pricing is quote-based; Activation, DAM, Supplier Data Manager and PricingHUB modules can change total cost materially. (-) The PricingHUB integration is recent (June 2026); check how pricing and product data will be unified in the core platform. (-) Native support for GS1/GDSN, ETIM or BMEcat was not confirmed on Akeneo's own pages; verify connectors through the App Store or partners.
Best for: Mid-market and enterprise brands and retailers that want a SaaS PIM suite with supplier onboarding, DAM, syndication and governed AI agents from one vendor.

## Vendor: Apimio
URL: https://pimatlas.com/vendors/apimio/
Product: Apimio PIM | Category: SMB PIM | HQ: Richmond Hill, Canada | Founded: Not disclosed | Ownership: Private (Apimio Inc.)
Deployment: SaaS | Segments: SMB, Mid-market | Standards: None verified
Pricing model: Published monthly plans with no annual contract: Basic PIM USD 199 (10,000 SKUs, 1 store, 3 users), Advanced PIM USD 399 (50,000 SKUs, extra stores from USD 50 to 100 each), Enterprise from USD 999 (custom). Brand Portal add-on USD 100 per month. 14-day free trial.

Apimio is a Shopify-native PIM and catalogue operations platform from Apimio Inc. for merchants that run one or several Shopify stores. It suits small and mid-sized D2C brands and wholesalers that want product data, supplier imports, data quality checks and feeds without an enterprise PIM project. Its focus on Shopify, monthly published pricing and self-serve onboarding distinguish it from general-purpose PIMs.

Apimio describes itself as a PIM built exclusively for Shopify. The product is organised into modules: Catalog Hub for central product data and multi-store sync, Quality Guard for data quality rules, Sale Scheduler, Supplier Bridge for supplier imports, Trade Portal for B2B dealer catalogues and Apimio AI for content generation. Scheduled imports, product feeds for Google and Meta, and wholesale line sheets complete the toolkit.

Pricing is public and monthly. Basic PIM costs USD 199 per month for up to 10,000 SKUs and one store, Advanced PIM costs USD 399 per month for up to 50,000 SKUs with additional stores priced on a sliding scale, and Enterprise starts at USD 999 per month. SKUs are counted once regardless of the number of languages or stores. Billing runs through Stripe or the Shopify invoice, with no annual commitment.

The Shopify App Store listing gives the developer address as Richmond Hill, Ontario, and dates the app to February 2021; Apimio's terms are governed by Delaware law. In September 2026 Apimio described a REST API, MCP server and agent-ready feed for AI shopping agents on its Advanced and Enterprise plans. Buyers outside Shopify, or those needing ERP connectors, GS1 data or B2B classification standards, should look at broader platforms.

Strengths: (+) Built specifically for Shopify, including multi-store and Shopify Markets workflows (+) Transparent monthly pricing with no annual contract and a 14-day trial (+) SKU counting deduplicated across languages and stores (+) AI content generation, supplier import mapping and data quality fixes in one tool (+) REST API, MCP server and agent-ready feed on Advanced and Enterprise plans
Considerations: (-) Works only with Shopify; Magento and WooCommerce are supported solely as migration sources (-) No disclosed ERP connectors, GS1/GDSN support or B2B classification standards (-) Very few public reviews on the Shopify App Store at the time of verification (-) No disclosed European presence or EU data hosting commitments
Best for: Shopify merchants with one or more stores and up to tens of thousands of SKUs that want a self-serve PIM with supplier imports, feeds and AI content at a fixed monthly price.

## Vendor: ATAMYA
URL: https://pimatlas.com/vendors/atamya/
Product: ATAMYA Product Cloud (successor to the eggheads Suite) | Category: Mid-market PIM | HQ: Bochum, Germany | Founded: 1990 | Ownership: Private (ATAMYA GmbH, formerly eggheads GmbH); no external investors disclosed
Deployment: SaaS | Segments: Mid-market, Enterprise, SMB | Standards: BMEcat, ETIM, ECLASS, UNSPSC, BIM
Pricing model: Four editions (Basic, Professional, Advanced, Ultimate) with limits on users, data domains and channels and unlimited SKUs. Basic is listed from EUR 599 per month; higher editions are quote-based, with paid add-ons.

ATAMYA (formerly eggheads) is a Bochum-based PIM vendor whose cloud-native ATAMYA Product Cloud replaces the earlier eggheads Suite. Built on MACH principles, it offers multi-domain PIM, syndication and agentic AI for data onboarding and enrichment, with a choice of large language models. Published edition pricing starting at EUR 599 per month and support for BMEcat, ETIM and ECLASS suit DACH manufacturers and wholesalers that want a SaaS PIM with transparent entry pricing.

ATAMYA GmbH, formerly eggheads GmbH, is registered in Bochum. It was founded in 1990 and traded as eggheads for most of its history, selling the eggheads Suite PIM. The company introduced the ATAMYA brand with its new cloud product in 2024, announced the rebrand of the whole company on 2 April 2025 to reflect a cloud-first strategy, and completed the change of legal name in May 2025. It has an office in Munich and a Swiss subsidiary in Zurich.

ATAMYA Product Cloud is a cloud-native, API-first and headless platform built on MACH principles. It covers product data management, multi-domain data, syndication, content enrichment, user and workflow management with a BPMN engine, and a marketplace of connectors. ERP connectors for SAP and Microsoft Dynamics are listed, alongside commerce connectors for Shopware, commercetools and Emporix and DAM links to pixx.io and Cloudinary. Classification support covers BMEcat, ETIM, ECLASS and UNSPSC.

The AI layer, branded Agentic PIM, is the main differentiator: a Smart Import Agent for supplier data onboarding, content generation, translation and quality automation, orchestrated in workflows with optional human review. In February 2026 partner netcare published a load test scaling the platform from 100,000 to 10 million SKUs. Forrester named ATAMYA a Notable Vendor in its Product Information Management Systems Landscape, Q3 2026. Customer references include Wilo, CHRIST, Steiff and Baier & Schneider.

Strengths: (+) Transparent entry pricing, with the Basic edition published from EUR 599 per month and unlimited SKUs in every edition (+) Agentic AI for supplier data onboarding with model choice across Anthropic, OpenAI, Google and self-hosted models (+) Support for BMEcat, ETIM, ECLASS and UNSPSC, relevant to technical wholesale and manufacturing (+) Listed connectors for SAP, Microsoft Dynamics, Shopware and commercetools (+) Long operating history since 1990 with a recently rebuilt cloud-native platform
Considerations: (-) Customers on the legacy eggheads Suite should clarify migration path, support horizon and costs for moving to ATAMYA Product Cloud (-) On-premise deployment and hosting region are not stated on the vendor website (-) The MCP server for external AI tools is still under development (-) Lower editions cap users, data domains and channels (Basic allows one user and one channel)
Best for: DACH manufacturers and wholesalers that want a cloud-native PIM with agentic supplier onboarding, B2B classification standards and published entry pricing.

## Vendor: AtroCore
URL: https://pimatlas.com/vendors/atropim/
Product: AtroPIM | Category: Open source PIM | HQ: Regensburg, Germany | Founded: 2020 | Ownership: Private, owner-managed (AtroCore GmbH)
Deployment: On-premise, SaaS | Segments: Mid-market, SMB | Standards: BMEcat, ETIM
Pricing model: Open-source edition free to self-host. SaaS Edition from EUR 400 per month (Bronze support included); SaaS Premium Edition from EUR 1,930 per month with all premium modules. Premium modules from EUR 20 per month on SaaS or one-off on-premise licences from EUR 500 plus 20% yearly updates. ERP integrations from EUR 1,900, e-commerce integrations from EUR 3,900.

AtroPIM is a GPLv3 open source PIM built on the AtroCore data platform by AtroCore GmbH of Regensburg, Germany. It suits mid-sized manufacturers and wholesalers, especially in the DACH region, that want a free self-hosted core and the option of paid SaaS hosting. Its distinguishing feature is a published, modular price list: premium modules such as ETIM, BMEcat, workflows and AI are bought individually.

AtroPIM is the product information management application of AtroCore, a configurable data management and integration platform. The source code is published on GitHub under the GNU GPLv3 licence, and the company describes itself as an owner-managed software business founded in 2020 with more than 20 staff. AtroCore reports over 300 users worldwide, about half of them in Germany, with manufacturers making up about 60% of the base and wholesalers and retailers about 20% each.

The commercial model is unusually transparent. The open-source edition runs on the buyer's own infrastructure at no licence cost, while the SaaS Edition starts at EUR 400 per month and the SaaS Premium Edition at EUR 1,930 per month. Functions that many buyers consider core, including workflows, versioning, data quality, translations, ETIM classification, the BMEcat adapter and the AI Integration module, are sold as individually priced premium modules. Prebuilt connectors for ERP and e-commerce systems such as SAP S/4HANA, Business Central, Shopware and Adobe Commerce are charged separately.

AtroCore 2.0, released in June 2025, added reusable filters, attributes that vary by product type, channel and region, a redesigned product page and faster imports and exports. Later 2025 updates introduced channel-specific product listings, completeness tracking by product classification, S3-compatible storage and finer role permissions.

Strengths: (+) Core code is open source under GPLv3 and can be self-hosted without licence fees (+) Public price list for SaaS hosting, support packages and every premium module (+) Native ETIM classification and BMEcat adapter modules for technical B2B catalogues (+) Underlying AtroCore platform can model non-product master data such as suppliers (+) Print output through a PDF Generator and an EasyCatalog adapter for Adobe InDesign
Considerations: (-) Workflows, versioning, data quality and AI are paid modules, so total cost depends on the module mix (-) Small vendor with 20+ staff, so check partner coverage and support capacity in your region (-) Each ERP or e-commerce connector carries a separate one-off fee (-) No verified GS1/GDSN data pool connectivity for consumer goods syndication
Best for: Mid-sized European manufacturers and wholesalers with technical catalogues that want an open source PIM with ETIM and BMEcat support and predictable published pricing.

## Vendor: Bluestone PIM
URL: https://pimatlas.com/vendors/bluestone-pim/
Product: Bluestone PIM | Category: Enterprise PIM | HQ: Tønsberg, Norway | Founded: 2016 | Ownership: Private, venture backed (2024 round with Scale Leap Capital and Narwhal Investments)
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1, ETIM, EPREL, NOBB
Pricing model: Quote-based usage pricing driven by users, SKUs and languages; no published prices. Contract terms are typically 3 to 12 months and marketplace apps are priced on the same basis.

Bluestone PIM is a Norwegian SaaS PIM built on a headless, API-first MACH architecture, aimed at mid-market and enterprise retailers, brands and manufacturers running composable commerce stacks. It distinguishes itself through an agent-oriented AI layer, an MCP gateway for external AI agents and a bring-your-own-model policy. Gartner named it a Visionary in the October 2026 Magic Quadrant for Product Information Management Solutions.

Bluestone PIM was founded in Norway in 2016 and became the first PIM vendor certified by the MACH Alliance in 2021. The platform is multi-tenant SaaS hosted on AWS, with sites in Frankfurt, Dublin, Singapore and the USA. The vendor states that every function sits behind an API, with more than 700 task-level endpoints documented in OpenAPI specifications, which suits buyers assembling a composable commerce stack.

The product covers core PIM, light DAM, workflow and an Extension Hub of pre-built connectors to commerce engines such as commercetools, Shopify, Adobe Commerce and BigCommerce, syndication tools such as 1WorldSync and Productsup, and partner extensions for GS1, ETIM, EPREL energy labels and the Norwegian NOBB building products database. Digital Product Passport work runs through partners Kolla and Traced Systems.

Since 2025 the roadmap has centred on AI. Named features include AI Enrich, AI Linguist, AI Analyst, AI DAM, AI Workflow and a conversational AI Agent, and an MCP gateway exposes catalogue data to external agents. The MACH Alliance recognised Bluestone as Agent Ready in 2026, and Gartner placed it as a Visionary in its 2026 Magic Quadrant. Pricing is usage-based and quote-only, with short contract terms.

Strengths: (+) API-first, headless design with more than 700 documented endpoints and MACH Alliance certification since 2021. (+) Broad shipped AI feature set, including a conversational AI Agent, an MCP gateway and bring-your-own-model support. (+) Named a Visionary in the 2026 Gartner Magic Quadrant for Product Information Management Solutions. (+) Flexible commercial terms: usage-based pricing and contracts typically between 3 and 12 months. (+) European footprint with five European entities and EU hosting in Frankfurt and Dublin.
Considerations: (-) GS1 and ETIM support is delivered through partner extensions (Qmica, Avensia); buyers needing native standards handling should test these in a demo. (-) BMEcat output is not listed among the extensions, which matters for German-speaking B2B distribution channels. (-) The headless model assumes integration capacity; teams without developers or a systems integrator should budget for implementation partners. (-) No public price list, so total cost must be modelled from a quote.
Best for: Mid-market and enterprise retailers and brands building a composable, API-driven commerce stack who want agent-ready product data.

## Vendor: Catsy
URL: https://pimatlas.com/vendors/catsy/
Product: Catsy PIM and DAM | Category: Mid-market PIM | HQ: Chicago, United States | Founded: Not disclosed (third-party profiles give conflicting years) | Ownership: Private; founded by Ceejay Teku, no external investors disclosed
Deployment: SaaS | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Quote-based. Packages for Shopify sellers (Essential, Multi-Store, Enterprise), industrial, distributor and BigCommerce use cases, plus a build-your-own option; no prices are published. A free trial is offered on the Essential package.

Catsy is a cloud PIM and DAM platform from Chicago for brands, manufacturers and distributors that sell through North American retailers, distributors and e-commerce platforms. It suits mid-market and smaller teams that want product data, digital assets and retailer-ready exports in one tool with guided onboarding. Its distinguishing feature is a library of native retailer and distributor templates, from Home Depot and Lowe's to Ferguson and Fastenal.

Catsy combines product information management and digital asset management in a single SaaS application. Core features include configurable attributes, bulk editing and validation, completeness and readiness dashboards, automated workflows, role-based access, version history and a self-serve portal for partners. Output options include marketplace channels, brand portals, product sheets and Shopify CSV exports, with a content delivery network for assets.

The integration catalogue is weighted towards North American trade. Retailer connectors include Amazon, Home Depot, Lowe's, Menards, Walmart, Wayfair, Costco and Target; distributor connectors include Fastenal, Ferguson, Graybar and Google Manufacturer Center. E-commerce connectors cover Shopify, BigCommerce, Adobe Commerce, WooCommerce, PrestaShop and Shopware, and ERP connectors include Epicor, Infor, Microsoft Dynamics, NetSuite, Sage and SAP.

Catsy is privately held and led by founder Ceejay Teku, with offices in Chicago and Brooklyn. Packaging is organised by use case, with Shopify, industrial, distributor and BigCommerce packages and tailored onboarding of 30 days, 90 days or a custom period. In 2026 Catsy published material on an MCP server for AI access to PIM and DAM data. European buyers should check data residency, local language support and coverage of European retailers before shortlisting.

Strengths: (+) PIM and DAM in one SaaS application with completeness and readiness dashboards (+) Native templates and connectors for major North American retailers and distributors (+) Broad e-commerce and ERP connector list, including Shopify, BigCommerce, NetSuite, Epicor and SAP (+) Use-case packages with tailored onboarding and managed delivery options (+) MCP server for connecting AI systems to governed PIM and DAM data
Considerations: (-) No European office, hosting region or GS1/GDSN support is documented (-) Prices are not published; package feature boundaries require a sales conversation (-) Named AI features beyond the MCP server are not documented on the website (-) Company facts such as founding year, headcount and funding are not disclosed
Best for: North American brands, manufacturers and distributors that need PIM and DAM with ready-made retailer and distributor exports.

## Vendor: censhare
URL: https://pimatlas.com/vendors/censhare/
Product: censhare (PIM, DAM and content platform) | Category: DAM and PIM suite | HQ: Munich, Germany | Founded: 2001 (first product released 2002) | Ownership: censhare GmbH is part of Entirely, the marketing technology group owned by the family-owned DuMont media group; acquisition date not verified
Deployment: SaaS, On-premise | Segments: Enterprise | Standards: None verified
Pricing model: Quote-based; offered as licence and rental (subscription) models according to company history, with no public prices.

censhare is a Munich-based content platform that combines DAM (digital asset management), PIM and content management in one system, with strong cross-media and print publishing through Adobe InDesign. It suits enterprises in retail, consumer goods and media whose main challenge is managing assets and marketing content alongside product data. censhare is part of Entirely, the marketing technology group of the DuMont media group, and offers censhare Cloud hosted in Germany as well as on-premise deployment.

censhare traces its origins to 2001, when development began on what became the CoWare Server, released in 2002. The company was renamed censhare AG in 2008 and converted to censhare GmbH in October 2020. It is headquartered in Munich, where development is based, and has offices in the UK, the US, Switzerland, France, Benelux, the Nordics and India. censhare is now one of the companies of Entirely, a marketing technology group that describes itself as part of the family-owned DuMont group.

The platform is asset and content centric. One data model holds digital assets, product information and editorial content, which supports cross-media publishing to web, print and other channels. Integration with Adobe InDesign and InDesign Server has been part of the product since early versions, and censhare participates as an official partner in Adobe's InDesign Markup Language (IDML) development. Within Entirely, censhare integrates natively with MARMIND for marketing resource management.

Deployment comes in two forms: censhare Classic for on-premise installations and censhare Cloud, launched in January 2024 and hosted by censhare in Germany. Named customers include Dyson, IHG Hotels and Resorts, Rewe Group, Vitra and IKEA. Buyers looking mainly for product data governance, supplier onboarding or syndication should test those areas carefully, because censhare's centre of gravity is DAM and publishing.

Strengths: (+) One platform for DAM, PIM and content management with a shared data model. (+) Long-standing Adobe InDesign and InDesign Server integration for catalogue and print production. (+) censhare Cloud hosted in Germany, plus on-premise censhare Classic for organisations that need self-hosting. (+) Enterprise references across retail, consumer goods and media, including Dyson, Rewe Group, Vitra and IKEA. (+) Native MARMIND integration for marketing resource management within the Entirely group.
Considerations: (-) DAM and publishing are the core strengths; PIM-specific features such as supplier onboarding and retailer syndication need careful testing. (-) Publicly documented AI features are limited to auto-categorisation, sentiment analysis and image recognition; ask for the generative AI roadmap. (-) Support for GS1/GDSN, ETIM, BMEcat or ECLASS was not verified for this profile. (-) The censhare website could not be reviewed for this profile, so editions and pricing should be confirmed directly with the vendor.
Best for: Enterprises whose product content work is driven by assets, marketing content and print or cross-media publishing.

## Vendor: Comosoft
URL: https://pimatlas.com/vendors/comosoft-lago/
Product: COMOSOFT LAGO (current version LAGO TANA) | Category: DAM and PIM suite | HQ: Hamburg, Germany | Founded: 1994 | Ownership: Part of Constellation Software Inc. since 2022 (via the Eversfrank Group)
Deployment: SaaS, On-premise | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based. No public prices.

COMOSOFT LAGO is a multichannel planning, PIM and DAM system from Hamburg-based Comosoft, built for retailers and mail-order companies that produce high volumes of flyers, catalogues and promotional content. Its distinguishing feature is tight Adobe InDesign integration for SKU-based layouts, combined with campaign planning, proofing and digital channel output. Comosoft has been part of Constellation Software since 2022, and customers include Lidl, Lowe's and Bass Pro Shops.

Comosoft GmbH was founded in Hamburg in 1994 and develops all its software there. It became part of the Eversfrank Group, a German print and media group, and since early 2022 has belonged to Constellation Software Inc. (CSI), the Toronto-headquartered software holding company. Comosoft also operates in Valencia, Spain, and in Plano, Texas, and lists customers such as Lidl, Lowe's, Bass Pro Shops, Raja, Bader, Peter Hahn and Erwin Müller.

LAGO is a planning and production system for marketing collateral. It combines PIM for article, price and product data, DAM for media, a Whiteboard for campaign and page planning, and channel plug-ins for print via Adobe InDesign, e-commerce, e-paper, mobile, social media and hyperpersonalised marketing (DIM). Proofing, monitoring and history apps support approval workflows. LAGO can take data from ERP systems and prioritise high-margin or excess-stock items using sales and inventory data.

The current release, LAGO TANA, was announced in January 2024 with a refreshed interface, a web dashboard, a server-based Asset Processor that replaces Photoshop-based image processing, and support for Adobe InDesign and Photoshop 2024. Comosoft reports that LAGO is offered on-premise, as SaaS and in cloud versions. AI support for metadata, image descriptions and translations is integrated, and AI-based campaign forecasting is under development.

Strengths: (+) Deep Adobe InDesign integration for automated, SKU-based production of flyers, catalogues and circulars (+) Combines campaign planning, PIM, DAM, proofing and multichannel output for retail marketing (+) Long-standing references among large grocery, DIY and mail-order retailers in Europe and North America (+) Backed by Constellation Software, a long-term owner of vertical software businesses (+) ISO/IEC 27001:2022 certification at the Hamburg headquarters
Considerations: (-) Print and promotional planning focus makes it less suited to B2B manufacturers needing technical classification (-) No ETIM, ECLASS, BMEcat or GS1 support is documented (-) Named ERP and commerce connectors are not published; integration effort should be scoped (-) AI features are described at a high level, and campaign forecasting is still in development
Best for: Grocery, DIY and mail-order retailers that produce large volumes of promotional print and digital collateral from product data.

## Vendor: Contentserv (Centric Software)
URL: https://pimatlas.com/vendors/contentserv/
Product: Centric PXM (formerly Contentserv) | Category: Enterprise PIM | HQ: Rohrbach, Germany | Founded: 2000 | Ownership: Owned by Centric Software, a subsidiary of Dassault Systèmes; acquisition from Investcorp announced 25 February 2025 at an enterprise value of EUR 220 million. Contentserv is no longer sold as an independent brand.
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based; no public prices were verified for this profile.

Contentserv is a German PIM and PXM vendor founded in 2000 that Centric Software, a Dassault Systèmes subsidiary, agreed to acquire in February 2025 for an enterprise value of EUR 220 million. The product is now sold as Centric PXM, combining the former Contentserv PIM with Centric Shoppingfeed syndication and digital shelf analytics. It suits consumer goods and industrial manufacturers that want product content management connected to Centric's PLM portfolio.

Contentserv was founded in Germany in 2000 and grew into a mid-market and enterprise PIM and PXM vendor serving more than 1,600 customers in 90 countries. Investcorp held a majority stake from 2019, during which recurring revenue grew roughly sixfold, and in 2023 Contentserv acquired the French feed management company Shoppingfeed. On 25 February 2025 Dassault Systèmes announced that its subsidiary Centric Software had agreed to acquire Contentserv for an enterprise value of EUR 220 million.

Contentserv is no longer marketed as an independent brand. Its website states that Contentserv is now part of Centric Software, and the product line is sold as Centric PXM, described as comprising the former Contentserv platform plus syndication and digital shelf analytics. Centric presents two packages, PXM for FMCG and Retail and PXM for Industrial Manufacturing. Contentserv GmbH continues to operate from Rohrbach and Munich in Germany.

The strategic logic is an end-to-end product flow from development in Centric PLM to digital commerce through Centric PXM. Centric Shoppingfeed publishes product content to more than 1,000 channels, and a supplier onboarding portal handles brand and supplier data. Buyers evaluating Contentserv today are evaluating a Dassault Systèmes product line, so roadmap, licensing and support terms should be confirmed with Centric.

Strengths: (+) Backing of Dassault Systèmes through Centric Software, with a stated path from Centric PLM to Centric PXM. (+) Established customer base of more than 1,600 customers in 90 countries at acquisition. (+) Centric Shoppingfeed publishes product content to more than 1,000 channels. (+) Packages for both FMCG and retail and industrial manufacturing. (+) Strong German and French presence through Contentserv GmbH and Shoppingfeed.
Considerations: (-) The Contentserv brand has been absorbed into Centric PXM; confirm product roadmap and support commitments for existing Contentserv installations. (-) Deployment options, including any on-premise or private cloud availability, were not stated on current Centric PXM pages. (-) Support for ETIM, ECLASS, BMEcat or GS1/GDSN was not confirmed on current product pages. (-) Pricing is quote-based and packaging under Centric may differ from former Contentserv editions.
Best for: Consumer goods, fashion and industrial manufacturers that want PIM, feed syndication and digital shelf analytics linked to a PLM from the same vendor group.

## Vendor: crossbase
URL: https://pimatlas.com/vendors/crossbase/
Product: crossbase PIM and MAM | Category: Mid-market PIM | HQ: Böblingen, Germany | Founded: Not disclosed (vendor cites more than 25 years of operation) | Ownership: Private, owner-managed by its founders and partners
Deployment: SaaS, Private cloud, On-premise | Segments: Mid-market, Enterprise | Standards: ETIM, ECLASS, UNSPSC
Pricing model: Three editions (Standard, Professional, Enterprise) with named-user and capacity limits; on-premise, private cloud or SaaS. No public prices.

crossbase is a German PIM and media asset management system from crossbase mediasolution GmbH in Böblingen, aimed at manufacturers of technical and building products. It combines product data, media, translation management and fully automated print publishing, and offers identical functionality on-premise, in a private cloud or as SaaS. Support for ETIM, ECLASS and CAD/BIM output, plus ISO 27001 certification, makes it a fit for DACH manufacturers producing catalogues and technical documentation.

crossbase mediasolution GmbH is an owner-managed software company based in Böblingen, near Stuttgart, with further offices in Wolfurt, Austria, and Entzheim, France. Its founders, Thomas Kern (CEO) and Albrecht Bosler (CTO), still lead the company, and the vendor cites more than 25 years of focus on product communication. All development is carried out in-house, and customers receive two releases per year.

The crossbase system combines PIM and media asset management with modules for translation, e-commerce export, fully automated print publishing and sales information portals. The data model is freely configurable with classifiable attributes, inheritance, product views and relationship types such as spare parts, accessories and successors. It supports ETIM, ECLASS and UNSPSC classification, generates IFC files for BIM, and connects to CAD component catalogues through partners in the 4PACE group. Customers can run the same functionality on-premise, in a private cloud or as SaaS, and can switch between them.

References are mainly DACH manufacturers, including KARL STORZ, Schöck, HOPPE, Gebr. Kemper, FLEX, Schmersal, KTM and Meusburger. crossbase obtained ISO 27001 certification in early 2026 and was again named a TOP 100 innovator among German mid-sized companies in June 2025. Its AI features cover translation, text generation, image description, classification suggestions and data quality checks.

Strengths: (+) Identical functionality across on-premise, private cloud and SaaS, with the option to switch later (+) Fully automated print publishing alongside PIM and MAM in one system (+) ETIM, ECLASS and UNSPSC classification plus IFC and CAD output for building and technical products (+) Broad set of shipped AI functions, including DeepL translation with glossaries and ETIM classification suggestions (+) ISO 27001-certified information security management (2026)
Considerations: (-) Presence and references are concentrated in Germany, Austria, Switzerland and France (-) Named ERP connectors (SAP, Microsoft Dynamics) and BMEcat export are not documented on the website (-) Edition limits on named users and REST API interfaces should be checked against expected growth (-) Company size and founding year are not published
Best for: DACH manufacturers of technical, medical or building products that need PIM, media management and automated catalogue production in one system.

## Vendor: Crystallize
URL: https://pimatlas.com/vendors/crystallize/
Product: Crystallize (PIM and headless commerce) | Category: ERP and commerce-native PIM | HQ: Skien, Norway | Founded: 2017 | Ownership: Private; incubated by New Normal Group, now described as an independent New Normal Group company
Deployment: SaaS | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Published monthly plans per tenant: Particle free with pay-as-you-go metering, Atom USD 949 per month plus metering (5,000 items, 500 orders included), Crystal tailored for enterprises. Overage charged per item, order, bandwidth and API call.

Crystallize is a Norwegian headless commerce platform built around a PIM core, which it calls the Product Universe. It suits digital teams and agencies building custom storefronts, subscription or configurable-product businesses that want product data, content and commerce in one GraphQL API. Its distinguishing features are flexible product modelling with shapes, usage-based pricing and early support for AI agents through an MCP server.

Crystallize AS was founded in 2017 by brothers Bård and Aleksander Farstad and incubated by New Normal Group. It is headquartered in Skien, Norway, with offices in Oslo and San Francisco, and operates a remote-first team. Customer logos include Mercedes-Benz, KIA, Plantasjen and Sandqvist, and Nordic consultancies such as Columbus, twoday and Knowit appear among partners.

The platform combines PIM, rich content management and commerce in one data model. Product data is modelled with configurable shapes and multi-dimensional taxonomies, supports multilingual and multicurrency catalogues and can drive product configurators, subscriptions and bookable services. Everything is exposed through edge-cached GraphQL APIs, webhooks and a CLI, with frontend guides for Next.js, Astro, Remix, Svelte and Vue.

Pricing is published and usage-based. The free Particle plan is aimed at testing, Atom costs USD 949 per month plus metering for items, orders, bandwidth and API calls, and the Crystal enterprise plan adds three tenants, SSO and invoice billing. In April 2026 Crystallize released a remote MCP server that lets AI tools inspect schemas and query catalogues in read-only mode.

Strengths: (+) Single GraphQL API for product data, content and commerce in headless builds (+) Flexible product modelling for configurable, subscription and bookable products (+) Published, usage-based pricing with a free entry plan (+) Remote MCP server released in April 2026 for agent and AI coding access (+) Unlimited users and languages on all plans
Considerations: (-) Developer-led platform that needs frontend and integration resources to deliver value (-) Metered pricing means costs scale with items, orders and API traffic (-) No verified support for GS1/GDSN, ETIM or BMEcat (-) PIM governance features such as supplier onboarding and syndication are lighter than in dedicated PIMs
Best for: Digital-first brands and agencies in the Nordics and wider Europe building headless storefronts that want PIM, content and commerce in one API with usage-based pricing.

## Vendor: DynamicWeb (Truvio)
URL: https://pimatlas.com/vendors/dynamicweb-pim/
Product: DynamicWeb PIM | Category: ERP and commerce-native PIM | HQ: Aarhus, Denmark | Founded: 1999 | Ownership: Acquired by SignUp Software in February 2025; the combined group has operated as Truvio (Solna, Sweden) since January 2026. Ultimate owners not disclosed
Deployment: SaaS | Segments: Mid-market, SMB | Standards: GS1
Pricing model: Quote-based. PIM is sold stand-alone or in a suite with Commerce and CMS; no public prices.

DynamicWeb PIM is the product information module of DynamicWeb, a Danish B2B commerce suite that combines PIM, e-commerce and CMS on one cloud platform. It suits manufacturers and distributors running Microsoft Dynamics 365, particularly Business Central, who want PIM and B2B webshop from one vendor. Since 2025 DynamicWeb has been part of SignUp Software, now rebranded as Truvio.

DynamicWeb was founded in Aarhus, Denmark, in 1999 and built its business around commerce and content software for Microsoft Dynamics customers. In February 2025 SignUp Software, a Swedish Microsoft ISV, announced the acquisition of DynamicWeb, which then had around 1,500 customers and 175 partners. In January 2026 the group launched the Truvio brand, uniting SignUp Software (ExFlow, Axtension), SKSoft and DynamicWeb with more than 480 employees and 5,000 customers.

The PIM handles product data, variants, inheritance, grid editing and asset management, and publishes to e-commerce, print via DynamicWeb Publications and EasyCatalog, marketplaces through Channable and ChannelAdvisor, and data pools such as GS1 and Icecat. Its main differentiator is depth on the Microsoft stack: dedicated connectors cover Business Central, Finance and Operations, NAV and AX, and an accelerated PIM integration for Business Central was launched in November 2025.

The platform is cloud-based with version-less upgrades and can be deployed as stand-alone PIM or within the full suite. AI support includes an OpenAI-based assistant and built-in translation. Pricing is quote-based.

Strengths: (+) Deep Microsoft Dynamics integration, including a dedicated PIM for Business Central and connectors for Finance and Operations, NAV and AX. (+) PIM, B2B commerce and CMS on a single platform, which reduces integration work for B2B webshops. (+) Built-in print output through DynamicWeb Publications and an EasyCatalog link to Adobe InDesign. (+) Cloud delivery with version-less upgrades and autoscaling. (+) Backing of the Truvio group and its Dynamics partner channel.
Considerations: (-) The strongest fit is Microsoft-centric; buyers on SAP or other ERPs should validate connector depth. (-) ETIM and BMEcat support is not documented on the PIM product page. (-) AI features are less extensive than at PIM-specialist vendors and rely on an OpenAI integration. (-) The 2025 acquisition and 2026 Truvio rebrand may affect roadmap and branding; ask about product continuity commitments.
Best for: B2B manufacturers and distributors on Microsoft Dynamics 365 Business Central that want PIM and webshop from one vendor.

## Vendor: Equadis
URL: https://pimatlas.com/vendors/equadis/
Product: Equadis PIM, Product Data Syndication (PDS), AI Content Engine (ACE) and Digital Shelf Analytics (DSA) | Category: PXM and syndication | HQ: Carouge, Switzerland | Founded: 1998 | Ownership: Private; Cathay Capital became majority shareholder in June 2026, with existing shareholders and management reinvesting
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1/GDSN, ETIM
Pricing model: Quote-based; no public price list.

Equadis is a Swiss SaaS platform combining PIM, product data syndication with its own GS1-certified GDSN data pool, AI content extraction and digital shelf analytics. It suits consumer packaged goods, cosmetics and consumer health brands that must supply complete, compliant data to European retailers. Gartner named Equadis a Niche Player in its first Magic Quadrant for Product Information Management Solutions in October 2026.

Equadis was founded in 1998, and its name is a contraction of Equal-Distance, reflecting a neutral position between brands and distributors. The company is registered in Carouge near Geneva and runs subsidiaries in France, Portugal, Slovakia and Lebanon. It reports more than 100 employees, over 600 brands and retailers in its community, including Procter & Gamble, Unilever, Nestlé and L'Oréal, and more than 950 retail connectors.

The suite has four modules. Equadis PIM centralises and enriches product data with validation workflows and channel-specific product sheets. Product Data Syndication distributes data through GS1 GDSN, Excel and API formats mapped to each retailer's data model. The AI Content Engine extracts information from artwork and documents, and Digital Shelf Analytics monitors listings and reviews. Equadis operates its own GS1-certified GDSN data pool and holds ETIM partner status.

In June 2026 Cathay Capital acquired a majority stake to fund hiring, international expansion and AI development, ending a self-funded history. In October 2026 Equadis was positioned as a Niche Player in Gartner's first Magic Quadrant for Product Information Management Solutions, which assessed 16 vendors, reflecting its specialisation in CPG, cosmetics and consumer health.

Strengths: (+) Own GS1-certified GDSN data pool combined with PIM in one platform (+) More than 950 retailer connectors with a strong base among French and Benelux retailers (+) AI Content Engine extracts data from packaging artwork and regulatory documents (+) Named Niche Player in the 2026 Gartner Magic Quadrant for Product Information Management Solutions (+) Digital shelf analytics closes the loop between published data and retailer listings
Considerations: (-) Focus on CPG, cosmetics and consumer health may limit fit for complex technical B2B catalogues (-) Change of majority ownership in June 2026 warrants questions on roadmap and pricing continuity (-) Pricing is not published (-) Presence outside French-speaking Europe and the Benelux is smaller, so check local support
Best for: Consumer goods, cosmetics and consumer health brands selling to French, Swiss and Benelux retailers that need PIM, GDSN publication and retailer-specific syndication from one vendor.

## Vendor: Ergonode
URL: https://pimatlas.com/vendors/ergonode/
Product: Ergonode PIM | Category: Mid-market PIM | HQ: Kraków, Poland | Founded: 2018 | Ownership: Private, founder-led with private investor backing
Deployment: SaaS, Private cloud | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Published annual plans excluding VAT: Free (EUR 0, 1,000 products, 2 users), Advance from EUR 5,990 per year, Scale from EUR 19,990 per year (300,000 products, 40 users), Enterprise quote-based with private cloud hosting. Concierge onboarding from EUR 1,500.

Ergonode PIM is a SaaS product information management system from Kraków, Poland, aimed at mid-market retailers and brands selling through e-commerce and marketplaces. It suits teams that want a quick start, published pricing and AI enrichment included in every plan. The original open source edition was archived in 2023, so Ergonode now competes as a commercial SaaS PIM with a free entry tier.

Ergonode was founded in 2018 by Marcin Piwowarczyk, its CEO, and first launched as an open source PIM under the Open Software License 3.0. In late 2021 it became an independent company and moved to a SaaS model; the public backend repository was archived on 13 February 2023. Buyers comparing open source PIM options should therefore treat Ergonode as a commercial SaaS product.

The platform covers attribute and product family modelling, workflows, automation rules, built-in DAM and publishing to Shopify, Magento, PrestaShop and marketplaces such as Amazon, eBay, Allegro and Etsy, with GraphQL and media APIs. Pricing is public: a Free plan for up to 1,000 products, Advance from EUR 5,990 per year, Scale from EUR 19,990 per year and a quote-based Enterprise tier that adds private cloud hosting, unlimited users and a dedicated customer success manager.

AI is central to recent releases. Every plan receives monthly AI credits for description generation, translation, attribute completion from text or PDFs and image generation in AI Photo Studio, launched in July 2026. Advanced DAM features followed in April 2026. In October 2025 Ergonode announced a strategic investment from Marcin Grzymkowski, founder of eObuwie, alongside expansion into the US, Canada, the UK and wider EMEA.

Strengths: (+) Transparent annual pricing with a free tier for up to 1,000 products (+) AI generation, translation, attribute extraction from PDFs and image generation included on every plan (+) Built-in DAM extended with advanced asset features in April 2026 (+) Ready-made connectors for Shopify, Magento, PrestaShop and marketplaces including Allegro (+) GraphQL and media APIs with published rate limits per plan
Considerations: (-) No longer open source: the OSL 3.0 repository was archived in February 2023 (-) AI credits expire monthly and premium models consume credits at up to 20 times the base rate (-) No verified support for B2B standards such as ETIM, BMEcat or GS1/GDSN (-) Staging environments are only included from the Scale plan upwards
Best for: Mid-market retailers and D2C brands in Central and Northern Europe that sell through Shopify, Magento or marketplaces and want a SaaS PIM with AI enrichment and published pricing.

## Vendor: Gepard (Bintime)
URL: https://pimatlas.com/vendors/gepard/
Product: Gepard PIM | Category: PXM and syndication | HQ: Tallinn, Estonia | Founded: 2015 (as a separate platform; Bintime development roots from 2005) | Ownership: Private; developed by Bintime, led by founder and CEO Sergii Shvets
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1/GDSN, ETIM, BMEcat, PI Standard, EPREL
Pricing model: Freemium: free tier for up to 200 products and 2 outputs with basic AI; Premium is quote-based with no product limits, advanced AI agents and 300+ connections.

Gepard PIM is a SaaS product content platform from Bintime, registered in Tallinn, Estonia, that combines PIM, syndication to retailers and marketplaces, supplier content collection and EU compliance data. It suits consumer electronics and appliance brands and retailers that must publish rich, standardised content to many channels. Its distinguishing features are a free entry tier, AI agents for mapping and PDF digitisation, and connectors for EPREL, GPSR and GDSN.

Gepard grew out of Bintime, a software company whose first product data management project started in 2005. Gepard became a separate product content platform in 2015, moved to multi-tenant SaaS in 2021 and reports a team of more than 150 people. Named customers and partners include HP, Dell, GS1, Icecat, Rakuten, Migros, MediaMarkt and Elgiganten.

The platform covers enrichment, validation, standardisation and syndication of product data, with retailer-specific content for brands and content collection for retailers. A large part of its proposition is regulatory and standards data: Gepard lists connections to EPREL energy labels, GPSR safety information, PPWR packaging, WEEE, ECHA, ETIM, GDSN, BMEcat and the PI Standard. PDF catalogue digitisation and a mapping module help onboard unstructured supplier data.

AI features include the AI Product Data Orchestrator and the AI Mapping Agent, released in August 2025. Bintime gained ISO/IEC 27001:2022 certification for Gepard PIM security in September 2025. A freemium tier covers up to 200 products and two outputs, with Premium pricing quoted individually.

Strengths: (+) Strong coverage of EU compliance data sources such as EPREL, GPSR, PPWR and WEEE (+) Supports GDSN, ETIM, BMEcat and PI Standard formats for B2B and retail exchange (+) AI agents for mapping inconsistent supplier data and digitising PDF catalogues (+) Long experience with consumer electronics content through Bintime and Icecat relationships (+) ISO/IEC 27001:2022 certification for Gepard PIM security, announced in September 2025
Considerations: (-) Premium pricing is not published, and the free tier is limited to 200 products (-) Strongest references sit in electronics and appliances, so check fit for fashion or food (-) Positioned mainly around syndication and content, so test internal governance workflows in detail (-) Mixed public statements on company history and experience should be clarified in due diligence
Best for: Consumer electronics, appliance and lighting brands and retailers that need standardised, compliance-ready product content syndicated to European retailers and marketplaces.

## Vendor: Goaland
URL: https://pimatlas.com/vendors/goaland/
Product: Goaland PIM | Category: SMB PIM | HQ: Paris, France | Founded: 2016 | Ownership: Private, founder-led (Goaland France SAS)
Deployment: SaaS, On-premise | Segments: SMB, Mid-market, Enterprise | Standards: None verified
Pricing model: Published per-user SaaS pricing billed annually: Core USD 163, Business USD 188 and Pro USD 226 per user per month, with tiered item, attribute and API limits and paid add-ons for video, display and print. Enterprise (SaaS or on-premise) is quote-based; a 30-day free trial is offered. EUR pricing is also available.

Goaland PIM is a French SaaS PIM that pairs product data management with a marketing content factory for print catalogues, videos, display banners and marketplace content. It suits small and mid-sized brands, manufacturers and retailers that want PIM and automated marketing output in one tool with published per-user pricing. Founded in Paris in 2016, Goaland lists customers such as Michelin, Holcim, Yves Rocher and Castorama.

Goaland France SAS was registered in Paris in November 2016 and is led by its founders, CEO Nicolas Michel and CTO Etienne Sola. It is a small, founder-led company that invests in research and development with partners such as École Centrale de Lille. Its website lists customers including Michelin, Holcim, Orano, Yves Rocher, Castorama, Weber and Supermarché Match, along with agencies such as Havas and DDB.

Goaland PIM centralises, structures and enriches product data with validation workflows, user rights, calculated fields and real-time quality checks for criteria such as text length and image resolution. It links images, video and 3D files to product records and distributes data to sales channels with monitoring. Its differentiator is the Marketing Factory: extensions for print catalogues and datasheets built on Adobe InDesign Server, video and display banner generation, digital shelf monitoring and a supplier and partner portal.

Pricing is published, which is uncommon in this market. Core, Business and Pro plans are charged per user per month with limits on items, attributes and API calls, and the Enterprise plan adds on-premise deployment and unlimited volumes. Named connectors cover Shopify, PrestaShop, Adobe Commerce and Amazon. No B2B classification standards such as ETIM or GS1 are documented, and the company has not published dated product or corporate news since 2024.

Strengths: (+) Combines PIM with automated marketing output for print, video, display banners and marketplaces (+) Published per-user pricing with a 30-day free trial (+) Integrated AI assistant for descriptions, tone variations, consistency checks and translation (+) Supplier and partner portal and digital shelf monitoring included as extensions (+) Enterprise plan offers on-premise deployment alongside SaaS
Considerations: (-) Small vendor with a customer base concentrated in France (-) No ETIM, ECLASS, BMEcat or GS1 support is documented, limiting fit for technical B2B distribution (-) Named ERP connectors are not published; ERP integration should be scoped (-) Per-user pricing and item limits in lower plans can raise costs for larger teams and catalogues
Best for: Small and mid-sized French brands and retailers that want PIM and automated marketing content production in one tool with transparent pricing.

## Vendor: Informatica (Salesforce)
URL: https://pimatlas.com/vendors/informatica/
Product: Informatica MDM Product 360 | Category: Product MDM | HQ: Redwood City, United States | Founded: 1993 | Ownership: Wholly owned by Salesforce since 18 November 2025; the product line is marketed as Informatica from Salesforce
Deployment: SaaS, On-premise | Segments: Enterprise | Standards: None verified
Pricing model: Quote-based; MDM SaaS on IDMC is sold through Informatica's consumption-based licensing, and no public Product 360 prices were verified.

Informatica MDM Product 360 is the product information management offer of Informatica, a data management company that Salesforce acquired on 18 November 2025. Product 360 runs as a multi-tenant SaaS application on the Intelligent Data Management Cloud and connects product data with customer, supplier and location domains. It suits large enterprises that want PIM inside a broader MDM and data governance estate, particularly organisations already invested in Salesforce or Informatica.

Informatica was founded in 1993 and built its business on data integration, data quality and master data management. Salesforce completed its acquisition of Informatica on 18 November 2025, and the products are now marketed as Informatica from Salesforce. Salesforce states that Informatica's catalogue, integration, governance, quality and MDM services will feed a single data pipeline with MDM on Data 360, while integration capabilities are combined with MuleSoft.

Product 360 is the PIM application in the Informatica 360 family alongside Customer 360 and Supplier 360. The current offer, MDM SaaS Product 360, is described as a cloud-native, multi-tenant SaaS application on IDMC that creates golden product records through match and merge, applies validation and enrichment rules, routes approvals and syndicates content using preconfigured templates. Self-managed Product 360 releases remain documented on Informatica's documentation portal, so customers on older versions should map their path to SaaS.

AI development follows Informatica's CLAIRE brand. The CLAIRE Product Experience Agent, introduced with the Fall 2025 IDMC release, was in private preview in early 2026. The main open questions for buyers concern roadmap: how Product 360 will sit beside Salesforce Data 360 and Agentforce, and whether packaging or pricing changes under Salesforce ownership.

Strengths: (+) Product data mastered alongside customer, supplier and location domains on one MDM platform. (+) Mature data quality, match and merge and governance tooling from Informatica's core data management portfolio. (+) Backing of Salesforce, with stated plans to connect Informatica MDM to Data 360, MuleSoft and Agentforce. (+) Headless MDM exposes golden records through APIs and MCP servers for AI agents. (+) Product 360 positions traceability data for Digital Product Passports (origin, materials, lifecycle).
Considerations: (-) Roadmap uncertainty after the November 2025 Salesforce acquisition; ask how Product 360 will relate to Salesforce Data 360 and Agentforce. (-) The CLAIRE Product Experience Agent was in private preview as of January 2026, so generative enrichment may not be generally available. (-) Support for GS1/GDSN, ETIM, BMEcat and ECLASS was not confirmed on current product pages; verify connectors in a scripted demo. (-) An enterprise MDM programme in scope and cost; less suited to buyers who need a standalone marketing PIM.
Best for: Large enterprises that run Informatica or Salesforce data platforms and want product information governed as one domain of a multidomain MDM.

## Vendor: Inriver
URL: https://pimatlas.com/vendors/inriver/
Product: Inriver PIM | Category: Enterprise PIM | HQ: Malmö, Sweden | Founded: 2007 | Ownership: Private; current investors not disclosed on Inriver's own pages
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based SaaS subscription; no public list prices were verified for this profile.

Inriver is a Swedish SaaS PIM (product information management) vendor founded in 2007 and headquartered in Malmö, used by more than 1,600 brands, manufacturers and retailers. It suits manufacturers and brands with complex products that need enrichment, syndication, print output and digital shelf feedback in one platform. Gartner named Inriver a Leader in its 2026 Magic Quadrant for Product Information Management Solutions, and its Inspire AI layer now includes agents and an MCP server.

Inriver was founded in Sweden in 2007, initially to serve the fast-growing Scandinavian telecom sector, and now positions its SaaS platform for enterprise manufacturers and brands. Its global headquarters is in Malmö, with offices in Stockholm, Munich, Amsterdam, Chicago and Davao City. Rohit Goyal became chief executive in 2024. The company reports more than 1,600 customers, including Bosch, Würth, Kohler, Carhartt and New Balance.

The platform covers PIM with content onboarding, Syndicate Advance for channel distribution, Digital Shelf Analytics, Print and Publish through Adobe InDesign, and Brand Store. Integration work in 2026 focused on ERP and commerce: SAP certified Inriver's S/4HANA Cloud and SAP Commerce Cloud connectors as built on SAP Business Technology Platform in May 2026, and Inriver received Microsoft Certified Software designation for manufacturing AI.

AI ships through Inriver Inspire. The Spring 2026 release added the Enhance Agent, the Expression Assistant, AI onboarding of unstructured documents and a visual workflow builder that orchestrates agents with conditional branching. The Summer 2026 release extended the MCP server first released in September 2025 and added Signals and Projects for tracking content gaps. Gartner named Inriver a Leader in the 2026 Magic Quadrant for PIM Solutions.

Strengths: (+) Named a Leader in the 2026 Gartner Magic Quadrant for Product Information Management Solutions (report dated 12 October 2026). (+) SAP-certified connectors for S/4HANA Cloud and SAP Commerce Cloud, built on SAP BTP (May 2026). (+) Manufacturer focus with Print and Publish automation through Adobe InDesign for catalogues and data sheets. (+) Shipped agentic features: Enhance Agent, AI onboarding of unstructured files, visual workflow builder and an MCP server. (+) Strong European footprint with offices in Malmö, Stockholm, Munich and Amsterdam.
Considerations: (-) SaaS only; organisations requiring on-premise deployment need a different vendor. (-) Pricing is quote-based, and modules such as Syndicate Advance, Digital Shelf Analytics and Print and Publish add cost. (-) Native support for ETIM, ECLASS, BMEcat or GS1/GDSN was not confirmed on Inriver's own pages; verify for B2B and CPG use cases. (-) Ownership and investor information is not published on Inriver's website; ask about financial backing during due diligence.
Best for: Manufacturers and brands with complex, multi-market product ranges that want an enterprise SaaS PIM with SAP integration and catalogue output.

## Vendor: Kontainer
URL: https://pimatlas.com/vendors/kontainer/
Product: Kontainer DAM and PIM | Category: DAM and PIM suite | HQ: Copenhagen, Denmark | Founded: 2001 | Ownership: Private, founder-led (Nanna Bentel and Jesper Sandberg)
Deployment: SaaS | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Published monthly USD prices. DAM from USD 500 (Core) to USD 2,220 (Enterprise). PIM including DAM from USD 640 (Business, 100 SKUs) to USD 2,265 (Premium, 10,000 SKUs) with SKU add-ons; Enterprise PIM for 50,000+ SKUs is quote-based. One month's notice.

Kontainer is a Danish DAM and PIM suite for brands that want product data, images, video and consent records in one GDPR-focused SaaS platform. It suits mid-sized fashion, furniture and consumer brands in the Nordics and DACH with modest SKU counts and heavy visual content. Its distinguishing features are published monthly pricing, photo consent and rights tools and a long list of creative and e-commerce integrations.

Kontainer was founded in Copenhagen by Nanna Bentel and Jesper Sandberg, who describe the business as delivering simpler workflows since 2001, and it also has an office in Munich. The company reports more than 350 client brands, 10,000 weekly users and 45 million assets shared each week, with named customers including Ricola, Les Deux, Select Sport, Gehl and the Danish Red Cross.

The platform started as DAM and added a Product Hub PIM that imports, maintains and shares product data and assets with web shops, B2B portals and retailers' PIM systems. DAM functions include brand portals, download presets, video streaming with subtitles, licence expiry reminders and a photo consent tool. Tailored export formats for lookbooks, price lists and marketplace feeds were added in March 2026.

Pricing is published in USD. PIM bundles include DAM and are tiered by SKU volume, product families, variations and dimensions such as language and currency, which makes Kontainer best suited to catalogues of up to tens of thousands of SKUs. Kontainer has also published guidance on MCP and on connecting AI assistants to its REST API, and on 23 September 2026 released a DeepL translation integration for PIM and DAM text fields.

Strengths: (+) Combined DAM and PIM with published monthly prices and one-month notice (+) Consent, licence and rights management tools, including AI facial recognition for photo consent (+) Over 300 integrations spanning e-commerce, CMS, creative tools and other PIMs (+) European company with Copenhagen and Munich offices and a GDPR-centred proposition (+) DeepL translation integration that reuses customer glossaries
Considerations: (-) PIM tiers cap SKUs, product families and dimensions, so large or complex catalogues move to custom Enterprise pricing (-) No verified support for GS1/GDSN, BMEcat or ECLASS exchange (-) Proposition is DAM-led, so test PIM depth such as workflows and completeness scoring in demos (-) MCP connectivity is described as API-based, so confirm the exact AI integration scope
Best for: Nordic and DACH fashion, furniture and consumer brands with image-heavy catalogues of up to tens of thousands of SKUs that want DAM and PIM from one European vendor at published prices.

## Vendor: novomind
URL: https://pimatlas.com/vendors/novomind-ipim/
Product: novomind iPIM | Category: Enterprise PIM | HQ: Hamburg, Germany | Founded: 2000 | Ownership: Private, owner-managed
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: ETIM
Pricing model: Quote-based SaaS subscription. No public prices.

novomind iPIM is a modular commerce PIM from Hamburg-based novomind, built for retailers, wholesalers and brands handling large, supplier-fed assortments. It runs as SaaS in novomind's own German data centre and adds modules for supplier onboarding, print, analytics and data search. Customers include LG Electronics Deutschland and HolzLand, which manages more than 1.3 million items from over 250 suppliers. Forrester listed novomind in its Q3 2026 PIM Systems Landscape.

novomind was founded in Hamburg in 2000 and remains owner-managed. The company reported revenue of EUR 57.7 million for 2024 and employs more than 500 people, around 480 of them in Hamburg and German branches. Its portfolio spans digital commerce (iPIM, iSHOP, iMARKET), customer service software (iAGENT) and the novomind AI Platform, so iPIM is often bought alongside other novomind products.

iPIM is a modular PIM system positioned for product experience management and product-related master data. Add-on modules cover supplier data onboarding (iPIM Supply), print publishing (iPIM Print), data quality reporting (iPIM Analytics) and product data search and download for partners (iPIM Buy). The platform supports multi-shop and multi-country set-ups and freely definable classification hierarchies, and novomind is a member of ETIM Deutschland. Operation is SaaS from novomind's own data centre in Germany, with ISO/IEC 27001 and TISAX certifications held at company level.

Reference customers are concentrated in retail and DIY. HolzLand uses iPIM for more than 1.3 million item records from over 250 industrial suppliers, and Transa processes around 90 supplier catalogues through iPIM Supply. In 2026 novomind announced projects with LG Electronics Deutschland and EFB-Elektronik, and Forrester included novomind among 23 vendors in The Product Information Management Systems Landscape, Q3 2026.

Strengths: (+) Supplier onboarding module (iPIM Supply) proven at high volume, such as more than 1,000 supplier catalogues at HolzLand (+) SaaS operated from novomind's own data centre in Germany, with ISO/IEC 27001 and TISAX certifications (+) Shipped AI smart mapping and AI text generation, backed by the separate novomind AI Platform (+) Single-vendor option for PIM, shop (iSHOP) and marketplace connectivity (iMARKET) (+) Financially stable, owner-managed vendor with EUR 57.7 million revenue in 2024
Considerations: (-) Named ERP connectors (SAP, Microsoft Dynamics) and commerce platforms outside the novomind stack are not documented publicly (-) BMEcat, ECLASS and GS1 support is not stated on the product pages and should be confirmed (-) On-premise deployment is not offered on the vendor website (-) International presence outside Germany is focused on MEA and APAC, with limited documented partner coverage elsewhere in Europe
Best for: German-speaking retailers, DIY chains and brand manufacturers that onboard large volumes of supplier data and want a German-hosted SaaS PIM.

## Vendor: Oro, Inc.
URL: https://pimatlas.com/vendors/orocommerce-pim/
Product: OroCommerce (Catalog Management and PIM) | Category: ERP and commerce-native PIM | HQ: Los Angeles, United States | Founded: Not disclosed | Ownership: Private; founded by Yoav Kutner, Jary Carter and Dima Soroka
Deployment: SaaS, On-premise | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based. Community edition source code is available under the Open Software License 3.0; Enterprise features, OroCloud hosting and OroIQ AI tools are part of commercial Enterprise licences.

OroCommerce is an open source B2B commerce platform from Oro, Inc. with native catalogue management and PIM functions. It suits manufacturers and distributors that want product data, pricing, ordering and buyer portals in one system rather than buying a separate PIM. Its product data model handles unlimited attributes, variants, kits and units of measure, and connectors to Akeneo, Pimcore or Salsify are available where a dedicated PIM is preferred.

Oro, Inc. was founded by the original creators of Magento and is headquartered in Los Angeles. Its main product, OroCommerce, is a B2B digital commerce platform whose core packages are licensed under the Open Software License 3.0, with commercial Enterprise editions and OroCloud hosting. The current release line is 7.0, with 7.0.4 published in August 2026 alongside the 6.1 long-term support branch.

The built-in PIM layer stores products with unlimited attributes and variants, multi-category assignment, localised content per language or region, units of measure, kits and configurable products, and digital assets such as images and spec sheets. Workflows can automate catalogue approvals and product launches. For organisations that already run a dedicated PIM, Oro offers an Akeneo connector and documents integration with Pimcore and Salsify.

AI functions are concentrated in commerce and back-office use cases. AI content generation for product descriptions uses OpenAI or Google Vertex AI, and the OroIQ assistants answer reporting questions and draft quotes and orders. Oro was recognised in the 2025 Gartner Magic Quadrant for Digital Commerce, so buyers should evaluate it as a commerce platform with PIM capabilities, with depth of syndication and supplier onboarding tested against standalone PIM vendors.

Strengths: (+) Product data, B2B pricing, quoting and ordering managed in one platform (+) Open Software License 3.0 core with a self-hosting option (+) Flexible B2B product model covering kits, configurables, units of measure and multi-category placement (+) AI content generation through OpenAI or Google Vertex AI plus OroIQ back-office assistants (+) Akeneo connector available for organisations that keep a dedicated PIM
Considerations: (-) PIM is a module of a commerce platform, so syndication to marketplaces and retailers is limited compared with standalone PIM (-) No verified native support for GS1/GDSN, ETIM or BMEcat (-) OroIQ and most AI tools are tied to the Enterprise licence (-) Pricing is not published and requires a sales quote
Best for: B2B manufacturers and distributors that want a single platform for product catalogue, customer-specific pricing and online ordering, and do not need extensive retailer syndication.

## Vendor: Perfion (Boyum IT Solutions)
URL: https://pimatlas.com/vendors/perfion/
Product: Perfion PIM | Category: Mid-market PIM | HQ: Denmark (city not disclosed) | Founded: 2002 | Ownership: Part of Boyum IT Solutions since 2023; Boyum was owned by Volpi Capital at the time of the acquisition
Deployment: SaaS | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Quote-based subscription with SaaS tiers (Starter, Growth, Business Pro) scaled by SKUs, asset storage, editors and languages, plus priced add-ons for ERP and e-commerce connectors, print publishing, Supplier Portal and Content Store.

Perfion PIM is a Danish product information management system owned by Boyum IT Solutions since 2023, aimed at mid-sized manufacturers, wholesalers and distributors running Microsoft Dynamics 365 or SAP Business One. Its distinguishing traits are a highly configurable data model, native ERP connectors and built-in print publishing, sold as tiered SaaS. AI features for content generation and import mapping arrived with release 2025-R3.

Perfion was founded in Denmark in 2002 and was acquired by Boyum IT Solutions, then a Volpi Capital portfolio company, in an agreement signed on 28 April 2023. At that point Perfion had about 40 employees across Denmark, the United States, Germany and the Netherlands. Boyum, a supply chain software group with more than 12,000 customers, positioned the deal around cross-selling into the Microsoft Dynamics and SAP Business One installed bases, and the product is now marketed as Perfion PIM by Boyum IT.

The platform targets manufacturers, distributors and wholesalers with complex, variant-heavy product data. Native connectors cover Dynamics 365 Business Central, Dynamics 365 Finance, SAP Business One and SAP S/4HANA, with e-commerce connectors for Shopify, Adobe Commerce, BigCommerce and Sana Commerce. Optional modules include DAM, print publishing, a Supplier Portal and the Content Store for sharing data with partners.

Releases now follow a versioned cadence. Perfion 2025-R3 (version 5.8.0, December 2025) introduced AI templates, GPT and API action commands, AI-assisted import mapping and a refreshed web client aligned with Boyum's design framework. Pricing is tiered SaaS on a quote basis, with SKU, editor and language allowances per tier.

Strengths: (+) Native connectors to Microsoft Dynamics 365 Business Central, Dynamics 365 Finance, SAP Business One and SAP S/4HANA. (+) Configurable data model suited to technical, variant-heavy catalogues in manufacturing and distribution. (+) Built-in print publishing and table design for catalogues and data sheets. (+) Backing of a larger ERP add-on group (Boyum IT) with an established partner channel. (+) Shipped AI features in 2025-R3, including generation templates and AI-assisted import mapping.
Considerations: (-) GS1, ETIM and BMEcat support is not documented on current vendor pages; buyers in regulated B2B channels should ask for evidence. (-) Reviewers note a learning curve when configuring the data model from scratch. (-) Several capabilities (DAM, Supplier Portal, connectors, print) are priced as add-ons, so total cost depends on the module mix. (-) On-premise availability is not stated on current pages; confirm deployment options during selection.
Best for: Mid-sized manufacturers and distributors on Microsoft Dynamics 365 or SAP Business One that need ERP-native PIM with print output.

## Vendor: Pimberly
URL: https://pimatlas.com/vendors/pimberly/
Product: Pimberly PIM | Category: Mid-market PIM | HQ: Manchester, United Kingdom | Founded: 2015 | Ownership: Private, founder-led (CEO Martin Balaam); GBP 5 million venture debt facility from Virgin Money (December 2025); equity shareholders not disclosed
Deployment: SaaS | Segments: Mid-market, Enterprise | Standards: None verified
Pricing model: Published annual subscriptions: Regular (EUR 36,000, 50,000 SKUs, 20 users), Pro (EUR 60,000, 100,000 SKUs, 50 users), Corporate (EUR 90,000, 250,000 SKUs, 100 users) and Enterprise (quote). Add-ons from EUR 7,500 a year for Pimberly AI to EUR 30,000 for Pimberly Catalog.

Pimberly is a UK SaaS PIM with integrated DAM, aimed at mid-market and enterprise retailers, distributors and manufacturers, with a strong position in UK retail and construction supply. It stands out for publishing annual prices in GBP, USD and EUR, a wide catalogue of commerce, ERP and marketplace integrations, and an AI add-on with bulk prompting and image attribute extraction. US revenue is growing quickly.

Pimberly was founded in Manchester in 2015 and is led by founder and CEO Martin Balaam. In its H1 2026 update the company reported annual recurring revenue of USD 10.4 million in June 2026, up 25% year on year, with 169 customers, 87 employees, 95% customer retention and a New York office. US sales doubled year on year and made up 28% of ARR. In December 2025 it secured a GBP 5 million venture debt facility from Virgin Money.

The platform combines PIM with DAM, workflows, a Vendor Portal for supplier onboarding, a Customer Portal, product data sheets and an eCatalog. The integrations page lists more than 20 e-commerce platforms, including Shopify, commercetools, Salesforce Commerce Cloud and Adobe Commerce, ERPs such as SAP, Dynamics 365, NetSuite and Sage, and marketplaces from Amazon to Walmart. Pimberly is an AWS Certified Solution Partner and a Shopify Certified Technology Partner.

Customers include JD Sports, Westcoast, Build-A-Bear, Harvey Nichols and Flight Centre. Memberships of the Construction Products Association, Builders Merchants Federation and Bathroom Manufacturers Association reflect a push into construction and building supply.

Strengths: (+) Transparent published pricing in GBP, USD and EUR, with SKU, user and storage allowances per tier. (+) Wide integration catalogue covering more than 20 e-commerce platforms, 15 ERPs and major marketplaces. (+) DAM, Vendor Portal and Customer Portal included in the platform. (+) Named AI tools (Pimbles, CopyAI, ColorAI, ImageAI) with a choice of LLM. (+) Disclosed operating metrics: USD 10.4 million ARR, 95% customer retention and 169 customers in 2026.
Considerations: (-) Entry pricing starts at EUR 36,000 a year, which is high for smaller catalogues. (-) GS1, ETIM and BMEcat support is not documented on the integrations page. (-) Customer base is concentrated in the UK and US; continental European references and partners are limited. (-) Enterprise DAM features and the AI module are priced as add-ons.
Best for: UK and US mid-market retailers, distributors and building supply manufacturers that want predictable published pricing and broad commerce integrations.

## Vendor: Pimcore
URL: https://pimatlas.com/vendors/pimcore/
Product: Pimcore Platform (PIM, MDM, DAM and DXP) | Category: Enterprise PIM | HQ: Salzburg, Austria | Founded: 2009 | Ownership: Majority owned by Nordwind Growth since October 2022; on 27 August 2026 Nordwind agreed to sell its majority stake to Tenzing, subject to antitrust approvals (pending at last verification). Co-founders remain co-CEOs.
Deployment: Private cloud, On-premise | Segments: Enterprise, Mid-market, SMB | Standards: None verified
Pricing model: Community Edition free under the Pimcore Open Core License for organisations with under EUR 5 million annual revenue, non-profits, academia and non-production use; Professional, Enterprise and Enterprise PaaS editions are licensed commercially on a quote basis.

Pimcore is an Austrian data and experience management platform that combines PIM, MDM, DAM and digital experience features on one open-core codebase. It suits manufacturers, retailers and wholesalers with development capacity that want a highly customisable, self-hosted or PaaS platform. Since release 2025.1 the Community Edition uses the Pimcore Open Core License instead of GPLv3. In August 2026 Tenzing agreed to acquire a majority stake, subject to antitrust approvals.

Pimcore was founded in 2009 in Salzburg, Austria, by Dietmar Rietsch and Matthias Blauth, who remain co-CEOs. The company reports more than 120,000 installations and over 400 enterprise customers, and Gartner included Pimcore in its 2026 Magic Quadrant for Master Data Management Solutions as well as its Digital Experience Platforms research. Nordwind Growth invested in October 2022. On 27 August 2026 Nordwind agreed to sell its majority stake to Tenzing, a London-based private equity investor, with completion subject to antitrust approvals.

The licence model changed in 2025. Pimcore Platform 2025.1 moved the Community Edition from GPLv3 to the Pimcore Open Core License (POCL), which keeps source code access but limits free use to organisations under EUR 5 million in annual revenue, non-profits, academia and non-production use. Commercial users license the Professional, Enterprise or Enterprise PaaS editions. Pimcore states that it is not a SaaS product: customers self-host, run it in a private cloud or use the managed PaaS.

The 2026 roadmap centres on the Data Spine and agentic AI. The Agent SDK, introduced in April 2026, builds agents into the core, and release 2026.3 added an Agent Bundle in beta, final MCP support, OAuth 2.0 for external agents and vector embeddings. Implementations usually run through certified partners and require development skills in PHP and Symfony.

Strengths: (+) Broad scope on one codebase: PIM, MDM, DAM, experience portals and CMS capabilities. (+) Deployment control: self-hosted on-premise, private cloud or managed Pimcore PaaS with multi-cloud and region options. (+) Free Community Edition for organisations under EUR 5 million revenue, with full source code access under POCL. (+) Included in the 2026 Gartner Magic Quadrant for Master Data Management Solutions. (+) Agentic AI foundations shipped in 2026: Agent SDK, MCP support, OAuth 2.0 for agents and vector embeddings.
Considerations: (-) Since 2025.1 the Community Edition is source-available under POCL; it is no longer OSI open source, and commercial use above the revenue threshold needs a paid edition. (-) Highly configurable platform; total cost depends heavily on implementation partner effort and in-house development capacity. (-) Pending majority stake sale to Tenzing (announced 27 August 2026); confirm roadmap and licensing continuity. (-) No multi-tenant SaaS option; buyers wanting a vendor-operated SaaS PIM should compare the PaaS service level carefully.
Best for: Manufacturers, retailers and wholesalers with development resources that want a customisable PIM, MDM and DAM platform under their own deployment control.

## Vendor: Pimics
URL: https://pimatlas.com/vendors/pimics/
Product: Pimics for Microsoft Dynamics 365 Business Central | Category: ERP and commerce-native PIM | HQ: Offices in Dresden, Brno and St. Gallen | Founded: 2001 (product in development since 2001, originally as Allium Katalog) | Ownership: Private (Allium group; German entity Allium (Deutschland) GmbH, Dresden)
Deployment: SaaS, On-premise | Segments: SMB, Mid-market | Standards: BMEcat, ETIM, ECLASS, GS1, GPC, DATANORM, UN/CEFACT
Pricing model: Monthly subscription with a one-year minimum term: Start EUR 390 (5,000 products), Lite EUR 690 (20,000), Professional EUR 990 (100,000), Corporate EUR 1,290 (unlimited). Users EUR 36, read-only users EUR 9, other Business Central users free; connectors such as BMEcat, DATANORM and Shopify are EUR 100 add-ons. Implementation is extra.

Pimics is a PIM extension that runs inside Microsoft Dynamics 365 Business Central, so product data, ERP data and users share one system. It suits small and mid-sized wholesalers and manufacturers on Business Central that need BMEcat, ETIM, ECLASS and GS1 support, supplier catalogue imports and catalogue publishing without a separate PIM platform. Published pricing starts at EUR 390 per month, and two major releases a year follow the Business Central cycle.

Pimics began in 2001 as Allium Katalog, a project to streamline printed catalogue preparation for Dynamics NAV 3.6. It was later renamed Pimics, extended into a full PIM module and certified by Microsoft for Dynamics from NAV 2015 onwards. The publisher operates as Allium, with a German entity, Allium (Deutschland) GmbH, in Dresden and further offices in Brno and St. Gallen. The vendor says the product is mainly aimed at small and medium businesses.

Because Pimics is a Business Central extension distributed through Microsoft AppSource, product and ERP data share the same server, user access and backup procedures. Functionality covers supplier data imports, classification, catalogue groups with inheritance, publications for print and e-shops, digital assets and checklists for data resolution. Standards support is broad for an ERP-native tool: BMEcat and DATANORM import and export, ETIM (including ETIM 10 from release 13.1), ECLASS, GS1 and GPC, and UN/CEFACT. Commerce connectors cover Shopify, Sana Commerce, Amazon and eBay.

Pimics ships two major releases a year in line with Business Central. In 2025 it added AI text generation, attribute suggestions, AI PDF import and DeepL translation, and reported customer growth of 52%. Release 13.1 (June 2026) added an upgraded Sana connector and ETIM 10 support, and release 14.0 (September 2026) added AI image descriptions and DeepL glossaries.

Strengths: (+) Runs natively inside Business Central, avoiding a separate PIM platform and integration layer (+) Broad B2B standards support: BMEcat, DATANORM, ETIM, ECLASS, GS1 and GPC (+) Transparent published pricing by product volume and user type (+) Shipped AI features including PDF supplier import, attribute suggestions, DeepL translation and image descriptions (+) Release cadence aligned with Business Central's two major releases per year
Considerations: (-) Only relevant to organisations running Microsoft Dynamics 365 Business Central or Dynamics NAV (-) Connectors and the AI Tool are paid add-ons, so total cost depends on the channel mix (-) Commerce connectors are limited mainly to Shopify, Sana Commerce, Amazon and eBay (-) Headquarters location and company size are not published
Best for: Small and mid-sized wholesalers and manufacturers on Microsoft Dynamics 365 Business Central that need B2B catalogue standards and supplier data imports inside their ERP.

## Vendor: Pimly
URL: https://pimatlas.com/vendors/pimly/
Product: Pimly on Salesforce | Category: ERP and commerce-native PIM | HQ: Evanston, United States | Founded: 2022 (per G2 and Crunchbase profiles) | Ownership: Private, venture backed (High Alpha led a USD 2.25 million strategic round in October 2025)
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Annual subscription. G2 lists a Pimly Starter Edition starting at USD 40,000 per year as provided by Pimly; higher editions are quote-based. Pimly's own website does not publish prices.

Pimly is a product information management (PIM) application built natively on Salesforce, from Evanston, Illinois. It suits manufacturers and B2B brands that run sales, service and commerce on Salesforce and want product data governed inside that platform. Its Product Intelligence layer grounds Salesforce Agentforce answers in governed product records, and Salesforce named Pimly a 2027 Partner of the Year in the ISV Transform and Scale category.

Pimly stores and governs product data inside the customer's Salesforce org, so product records are available in Sales, Service, Revenue and Commerce clouds without a separate synchronised system. CEO and founder Mike Dannenfeldt describes the product as a response to Salesforce customers asking for rich product information in their CRM. Pimly 4.0 (January 2025) added configurable workflows, permission and user groups, and linked items for bundles, compatibility and add-ons.

The focus since 2025 has been AI on Salesforce. Product Intelligence, listed on Salesforce AgentExchange in March 2026, positions governed product data as the knowledge layer for Agentforce. SmartSync, shipped in August 2026, sends changed product records to Agentforce Revenue Management, Agentforce Commerce, Sales Cloud, Service Cloud and Shopify, with each step logged. Pimly also offers DAM with expiring share links and an Asset Search feature that answers questions from attached documents.

High Alpha led a USD 2.25 million strategic round in October 2025, when Pimly reported almost 80 million product data operations a month. Named customers include GE Vernova, Cognex, Hormel Foods and Kraft Heinz. The architecture is a strong fit for Salesforce-centred organisations and a constraint for those whose commerce or CRM stack sits elsewhere.

Strengths: (+) Runs natively inside Salesforce, so product data is available across Sales, Service, Revenue and Commerce clouds without middleware (+) Product Intelligence grounds Agentforce responses in governed product records (+) SmartSync publishes only changed records to Salesforce clouds and Shopify with logged steps (+) Pimly 4.0 workflows, permission groups and linked items for bundles and compatibility (+) Named a Salesforce 2027 Partner of the Year (ISV, Transform and Scale)
Considerations: (-) Requires Salesforce; organisations on other CRM or commerce platforms gain less from the native architecture (-) Entry pricing listed on G2 starts at USD 40,000 per year, above many SMB PIM tools (-) Retailer syndication, GS1/GDSN and ETIM support are not documented (-) Small, recently funded vendor; check roadmap commitments and support coverage in Europe
Best for: Manufacturers and B2B brands that run sales, service and commerce on Salesforce and want product data governed inside that platform for reps and Agentforce agents.

## Vendor: PIMworks
URL: https://pimatlas.com/vendors/pimworks/
Product: PIMworks360 | Category: SMB PIM | HQ: Chennai, India | Founded: Not disclosed | Ownership: Private; product of Mobius Knowledge Services Pvt. Ltd., described by PIMworks as bootstrapped
Deployment: SaaS | Segments: SMB, Mid-market | Standards: None verified
Pricing model: Quote-based, priced on the number of SKUs and consumption, with e-commerce services charged by requirement. A 30-day free trial without a credit card is offered; no prices are published.

PIMworks360 is a SaaS product information management (PIM) platform from PIMworks, a product of Mobius Knowledge Services with offices in Chennai and New York. It suits small and mid-sized e-commerce sellers and brands that need to enrich catalogues and syndicate them to marketplaces such as Amazon, eBay and Walmart. Its distinguishing features are optional enrichment services from its parent company and the Digisense 360 digital shelf analytics add-on.

PIMworks360 provides a central product repository, automated content enrichment, multichannel syndication with templates and APIs, centralised digital asset management and collaborative workflows with hierarchy management and permissions. PIMworks states that the platform handles more than one million SKUs and exports to more than 20 marketplaces. Named connections include Amazon, eBay, Walmart and Houzz, plus Shopify, Adobe Commerce, BigCommerce, WooCommerce and the Acumatica ERP.

PIMworks was founded under the umbrella of Mobius Knowledge Services, an India-based content and data services company that cites more than 1,000 professionals and offices in Chennai, Coimbatore, New York and London. This lets PIMworks combine software with outsourced catalogue enrichment, which it charges according to requirement. A companion product, Digisense 360, covers price monitoring, rank tracking, marketplace SEO, review analysis and A+ content generation.

Public information about PIMworks is limited. The website does not disclose a founding year, product release history or named AI features beyond Enrich with AI, and no product announcements from 2024 to 2026 were found. Buyers should request a product roadmap, security documentation and reference customers in their region before shortlisting.

Strengths: (+) Marketplace-oriented syndication to Amazon, eBay, Walmart, Houzz and more than 20 marketplaces (+) Option to combine software with outsourced catalogue enrichment services from Mobius Knowledge Services (+) Digisense 360 add-on for price monitoring, rank tracking and review analysis (+) Connectors for Shopify, Adobe Commerce, BigCommerce and WooCommerce (+) 30-day free trial without a credit card
Considerations: (-) Limited public information on release history, AI features and security certifications (-) No European office, hosting region or GS1/GDSN support is documented for PIMworks (-) Prices are not published; costs depend on SKUs, consumption and services (-) No product announcements from 2024 to 2026 were found; ask for a current roadmap
Best for: Small and mid-sized e-commerce sellers focused on marketplaces that want PIM software combined with outsourced catalogue enrichment.

## Vendor: Plytix
URL: https://pimatlas.com/vendors/plytix/
Product: Plytix | Category: SMB PIM | HQ: Málaga, Spain | Founded: 2015 (per G2 seller profile) | Ownership: Private, venture backed (Crunchbase lists investors including Trestle Partners and an EU Horizon 2020 grant); founders Morten Poulsen (CEO) and Alejandro López (CTO)
Deployment: SaaS | Segments: SMB, Mid-market | Standards: None verified
Pricing model: Tiered plans priced by SKU count. The Plytix website lists Standard (free, no expiry), Pro starting at 149 per month and Enterprise starting at 699 per month (currency not stated); add-ons such as AI Unlimited, Feed Management, Brand Portals, Product Data Sheets and channel connectors cost 179 per month each. All plans include unlimited seats.

Plytix is a SaaS product information management (PIM) platform with built-in DAM and feed management, run mainly from Málaga, Spain. It suits small and mid-sized brands, manufacturers and retailers that sell through Shopify, marketplaces and retail partners and want transparent pricing. Its distinguishing features are a free-forever Standard plan, unlimited seats on every tier, published add-on prices and frequent releases of AI image and content tools.

Plytix combines a PIM, a digital asset manager and channel feeds in one SaaS application. Teams can model attributes and variants, track completeness, generate PDF product data sheets, publish brand portals for partners and build feeds for around 30 sales channels and marketplaces, including Google Shopping, Meta, Amazon and Zalando. Connectors cover Shopify, BigCommerce, WooCommerce, Adobe Commerce and PrestaShop, and an official Make app (September 2026) supports no-code automation.

Pricing is unusually transparent for the PIM market. The Standard plan is free with no trial expiry, Pro starts at 149 per month and Enterprise at 699 per month, with prices scaling by SKU count. Add-ons such as Feed Management, Brand Portals and AI Unlimited cost 179 per month each, and every plan includes unlimited seats and 500 AI credits a month.

Plytix AI generates and translates copy and flags missing data, and releases in 2026 concentrated on AI image tools and new channels such as ChatGPT feeds. Plytix states that it has been the top-ranked PIM on G2 for six consecutive years, and G2 listed it first among PIM tools in its Best Commerce Products 2024. Buyers with complex B2B data models, GS1/GDSN requirements or strict data residency needs should test those areas in a pilot, since they are not documented on the website.

Strengths: (+) Published pricing with a free-forever Standard plan and unlimited seats on every plan (+) PIM, DAM, feed management, brand portals and product data sheets in one application (+) Broad channel list, including Zalando, bol.com, Otto, Amazon, Google Shopping and ChatGPT feeds (+) Frequent releases, with AI image editing, relighting and upscaling shipped during 2026 (+) Ranked the top PIM in G2's Best Commerce Products 2024
Considerations: (-) Add-ons at 179 per month each can raise the total cost for multichannel use (-) No GS1/GDSN, ETIM or BMEcat support is documented (-) ERP integration is generic import or API; no named ERP connectors are listed (-) Currency and billing terms are not stated on the plan cards; confirm before budgeting
Best for: Small and mid-sized brands and retailers selling through Shopify and marketplaces that want an affordable PIM with DAM and feeds and published prices.

## Vendor: Precisely
URL: https://pimatlas.com/vendors/precisely-enterworks/
Product: Precisely EnterWorks | Category: Product MDM | HQ: Burlington, United States | Founded: 1968 (Precisely, originally Whitlow Computer Systems, later Syncsort) | Ownership: Private, private-equity backed (Clearlake Capital and TA Associates since 2021; minority investors include Centerbridge Partners, Insight Partners and Partners Group)
Deployment: On-premise, Private cloud | Segments: Enterprise, Mid-market | Standards: GS1/GDSN
Pricing model: Quote-based; Precisely does not publish EnterWorks prices.

Precisely EnterWorks is a multi-domain master data management (MDM) platform with built-in PIM, DAM, a supplier portal and GDSN syndication, sold by Precisely, the data integrity company based in Burlington, Massachusetts. It suits large manufacturers, distributors and retailers that want product, supplier and location data mastered in one model. Its distinguishing feature is the link to the wider Precisely portfolio, including the Data Integrity Suite, address validation and enrichment datasets.

EnterWorks is the product and multi-domain MDM application within the Precisely data integrity portfolio. It reached Precisely through the June 2021 agreement to acquire Winshuttle from Symphony Technology Group, which had merged EnterWorks into Winshuttle in 2019. Precisely positions EnterWorks for product, customer, supplier and location domains, with a configurable data model, validation, matching and survivorship rules for golden records, a Vendor Portal for supplier contributions and an integrated DAM.

Release 11.3 (October 2025) connected EnterWorks to the Data Governance service of the Precisely Data Integrity Suite, so mastered attributes can be catalogued and linked to glossary terms, policies and KPIs, and refreshed the user interface to match the suite. The GDSN Connector publishes to and subscribes from certified data pools such as 1WorldSync and Syndigo. Named customer stories include Fender, Johnstone Supply and Graco.

In October 2026 Precisely previewed the Precisely Platform, which brings master data management together with integration, quality, governance and location intelligence, with general availability expected in early 2027. Buyers should ask how EnterWorks maps onto that platform, which deployment models are offered (customer forums show both self-managed and Precisely-hosted installations) and which AI features are generally available in their version.

Strengths: (+) Multi-domain model covering product, supplier, customer and location data, with custom domains configurable without code (+) Built-in DAM, Vendor Portal for supplier onboarding and configurable taxonomies for large catalogues (+) GDSN Connector for publishing to and subscribing from certified data pools such as 1WorldSync and Syndigo (+) SAP-certified connector and integration with Precisely Data Integrity Suite governance, address validation and enrichment datasets (+) Named AI features for translation, description generation, image data extraction and auto-classification, plus MCP access for AI clients
Considerations: (-) The roadmap is tied to the Precisely Platform, in preview since October 2026; confirm the EnterWorks migration path and timelines (-) Precisely does not state deployment options on the EnterWorks product page; confirm SaaS, hosted and on-premise availability (-) Pricing is quote-based, and part of the value depends on adopting other Precisely modules (-) European implementation partners and hosting regions for EnterWorks are not disclosed
Best for: Large manufacturers, distributors and retailers that need product and supplier master data, GDSN syndication and data governance from one enterprise data vendor.

## Vendor: Productsup
URL: https://pimatlas.com/vendors/productsup/
Product: Productsup | Category: PXM and syndication | HQ: Berlin, Germany | Founded: 2010 | Ownership: Private; Nordwind Capital holds a majority stake, with Bregal Milestone leading a Series B of more than USD 70 million (EUR 65 million) in 2022
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based subscription; price depends on SKU count, number of feeds, platform modules, sync frequency, seats, countries and special requests. No prices are published.

Productsup is a product content syndication and feed management platform from Berlin, founded in 2010, that distributes product data to more than 2,500 advertising, marketplace, retailer and AI channels. It suits brands and retailers with large catalogues selling across many channels and countries. In 2026 it added AI channels, AI Enrich and AI Visibility to prepare catalogues for AI assistants and agentic commerce.

Productsup was founded in Berlin in 2010 and is led by co-founder and CEO Johannis Hatt. Nordwind Capital holds a majority stake, and Bregal Milestone led a Series B of more than USD 70 million in April 2022. The company reports more than 1,000 customers, around 200 employees and offices in Berlin, 's-Hertogenbosch, New York and Hyderabad.

The platform imports product data from PIM, ERP and commerce systems, transforms it with rules and optimisation, and publishes it to more than 2,500 channels, including Google, Meta, Amazon and TikTok. Productsup also supports retailer and distributor syndication and supplier onboarding. A February 2025 partnership with Syndigo gives Syndigo customers access to Productsup's EMEA channel network and Productsup customers access to Syndigo's North American retailer network.

In 2026 Productsup focused on agentic commerce. AI channels publish catalogues to OpenAI and Perplexity, AI Enrich generates conversational product content, and AI Visibility measures how products appear in AI assistants. Productsup became an approved Walmart Marketplace solution provider in September 2026. It is a syndication layer, so it relies on a PIM or ERP as the master source of product data.

Strengths: (+) More than 2,500 advertising, marketplace, retailer and AI channel integrations (+) AI channels, AI Enrich and AI Visibility launched in 2026 for AI assistant discovery (+) European base with offices in Berlin and 's-Hertogenbosch and EMEA marketplace coverage (+) Partnership with Syndigo that extends reach to North American retailers (+) Approved Walmart Marketplace solution provider covering listings, pricing and orders
Considerations: (-) Productsup is a syndication and feed layer; a PIM or ERP remains the master record for product data (-) Prices are not published and scale with SKUs, feeds, modules, sync frequency and countries (-) No GS1/GDSN data pool certification is documented on its website (-) AI Enrich and AI Visibility are new in 2026; check maturity and governance controls in a pilot
Best for: Brands and retailers with large catalogues that need to syndicate product content to many marketplaces, ad channels and AI assistants across countries.

## Vendor: Profisee
URL: https://pimatlas.com/vendors/profisee/
Product: Profisee MDM | Category: Product MDM | HQ: Alpharetta, United States | Founded: 2007 | Ownership: Private, private-equity backed (Pamlico Capital majority investment in 2021; ParkerGale Capital and management retain minority stakes)
Deployment: SaaS, Private cloud, On-premise | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based and volume-based: customers pay for the number of records mastered, with unlimited domains and attributes included and bulk pricing for larger volumes. Two editions are listed (Enterprise Edition for full MDM, Application Edition for reference data management) with SaaS or PaaS deployment; no prices are published.

Profisee MDM is a multi-domain master data management (MDM) platform from Profisee, based in Alpharetta, Georgia. It suits organisations on the Microsoft data stack that need product, customer or supplier golden records, since it runs as a native workload in Microsoft Fabric as well as SaaS, PaaS and on-premise. Gartner named Profisee a Leader in its 2026 Magic Quadrant for Master Data Management, and pricing is based on record volume with unlimited domains.

Profisee was founded in 2007 and has been majority owned by Pamlico Capital since a 2021 growth investment. Its platform covers matching, merging and golden record management (Enterprise Edition) and reference data management (Application Edition, positioned as an upgrade path from Microsoft Master Data Services). Product data is one of several domains, priced by record volume with unlimited domains and attributes.

The Microsoft relationship shapes much of the roadmap. In March 2026 Profisee made a Fabric-native MDM workload generally available, added Fabric Open Mirroring to write golden records to OneLake in Delta Lake format and released the Profisee MCP Server. The 2026.R2 release in July re-architected the Aisey AI agent, and in September 2026 Profisee Anywhere introduced management of master data directly in Fabric OneLake.

Gartner named Profisee a Leader in the 2026 Magic Quadrant for Master Data Management Solutions. The company reported 45 percent year-over-year growth in new sales for the first half of its fiscal 2027. Product information teams should note that Profisee is a general MDM platform: syndication to retailers, DAM and marketing enrichment are outside its documented scope and are normally handled by a PIM.

Strengths: (+) Native workload in Microsoft Fabric with Open Mirroring of golden records to OneLake (+) Broad deployment choice: SaaS, PaaS in Azure, AWS or Google Cloud, hybrid and on-premise (+) Record-volume pricing with unlimited domains and attributes (+) Aisey AI agent and Profisee MCP Server for two-way access from Copilot and other MCP clients (+) Named a Leader in the 2026 Gartner Magic Quadrant for Master Data Management Solutions
Considerations: (-) Strongest fit is a Microsoft-centred data estate; check connectors for other data platforms (-) No documented PIM features such as channel syndication, DAM or retailer templates (-) No product data standards such as GS1/GDSN are documented (-) Prices are not published, and the pricing page indicates up-front billing
Best for: Organisations standardised on Microsoft Azure and Fabric that need multi-domain master data, including product, with flexible deployment and volume-based pricing.

## Vendor: Propel Software
URL: https://pimatlas.com/vendors/propel/
Product: Propel PIM (with Propel PLM and QMS) | Category: Mid-market PIM | HQ: Redwood City, United States | Founded: 2015 (per G2 seller profile) | Ownership: Private; board members include partners from Norwest Venture Partners, Jackson Square Ventures, Cloud Apps Capital Partners and Ankona Capital
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Published annual licence ranges by tier: Small (PLM and QMS) USD 10K to 150K; Medium (PLM, QMS and PIM) USD 50K to 500K with USD 15K to 200K one-time setup; Large (PLM, QMS, PIM and PVM) USD 250K to 500K with USD 50K to 400K setup. Final pricing is quote-based by user role editions.

Propel PIM is the product information management module of Propel Software's product value management platform, which also includes PLM and QMS and runs on Salesforce. It suits manufacturers in medical devices, high tech and industrial equipment that want engineering, quality and commercial product data on one system. Its distinguishing feature is the direct handover from PLM records into PIM, supported by Propel One agentic AI and published pricing ranges.

Propel Software, based in Redwood City, California, sells a cloud platform built on Salesforce that combines PLM (product lifecycle management), QMS (quality management) and PIM. Propel PIM aggregates product data from multiple sources, including direct sharing from Propel PLM, and adds category hierarchies with channel-specific alternates, parent-child SKU variants, approval workflows, role-based permissions and a product digital asset manager. Content is distributed to channels by API or file export using saved templates.

AI is delivered through Propel One, an agentic AI suite on Salesforce Agentforce launched in March 2025, and Propel MCP, which reached production in June 2026. In August 2026 Propel launched Manufacturing Hub, which sends approved change data to ERP systems such as NetSuite, SAP, Oracle JD Edwards and Microsoft Dynamics. Propel reported 42 percent bookings growth for the fiscal year ended 31 January 2026, driven partly by migrations from Oracle Agile PLM.

Propel publishes pricing ranges by tier, and PIM is included from the Medium tier. Buyers whose priority is retail syndication should note that the PIM page does not document retailer templates, GDSN or marketplace connectors, and that Propel's product focus remains engineering-led manufacturers.

Strengths: (+) PIM, PLM and QMS on one platform, so engineering data flows into commercial product content without a separate integration (+) Propel One agentic AI on Salesforce Agentforce for content gap detection and copy generation (+) Production MCP support for Claude, ChatGPT and Microsoft Copilot (June 2026) (+) Published pricing ranges by tier, including one-time setup fees (+) Manufacturing Hub integration to ERP platforms including NetSuite, SAP, Oracle JD Edwards and Microsoft Dynamics
Considerations: (-) PIM is sold within PLM and QMS bundles; standalone PIM pricing is not published (-) No documented retailer templates, GS1/GDSN or marketplace connectors on the PIM page (-) European offices, partners and hosting regions are not disclosed (-) Built on Salesforce, so platform licensing and data residency follow Salesforce terms
Best for: Engineering-led manufacturers in medtech, high tech and industrial equipment that want PIM, PLM and QMS on one Salesforce-based platform.

## Vendor: Quable (Ibexa)
URL: https://pimatlas.com/vendors/quable/
Product: Quable PIM | Category: Mid-market PIM | HQ: Paris, France | Founded: 2013 | Ownership: Part of Ibexa group since 2025 (Quable states it joined Ibexa in 2025; its history page says 2020 to 2023). Group owners not disclosed
Deployment: SaaS | Segments: Mid-market, SMB | Standards: ETIM
Pricing model: Quote-based. Two plans: Standard (up to 50,000 products, seat-based) and Advanced (unlimited products and seats); both include unlimited channels. No public prices.

Quable PIM is a French SaaS PIM with built-in DAM, aimed at mid-market brands in fashion, beauty, food and retail that want a no-code tool with fast onboarding. Since 2025 it has been part of Ibexa, a European martech group combining DXP, CDP, customer engagement and PIM. Distinguishing points are an integrated DAM and brand portal, a Shopify Plus app and AI agents being built on the Ibexa platform.

Quable was founded in Paris in 2013 and launched its PIM in 2015, building the product with its first customers. The vendor claims more than 300 brands and over 40 million products under management, with logos including Petit Bateau, Lalique, Bollinger, Bioderma, Lanvin and Sisley. Quable states that it joined the Ibexa group in 2025, and in February 2026 Ibexa announced the unification of Quable, Actito, Qualifio, Raptor and a new Cohesivo CMS into one orchestration platform with more than 350 employees and 1,000 customers.

The product is a no-code SaaS PIM with a built-in DAM, brand portal, PDF product sheet builder, feeds, completeness dashboards and an app store. The Advanced plan adds visual workflows, smart calculated fields, automation rules and a custom app SDK. Integration is through REST and GraphQL APIs, webhooks and apps, including a Shopify Plus app. ETIM classification is supported for technical products.

AI text generation and translation are included in every plan. Agentic capabilities are in beta on the Ibexa Agentic Marketing Platform. Pricing is quote-based across two plans.

Strengths: (+) Integrated PIM, DAM and brand portal in one no-code SaaS product. (+) Strong reference base among French and European consumer brands in fashion, beauty and wine and spirits. (+) Unlimited channels on both plans, with REST, GraphQL and webhooks for integration. (+) Membership of the Ibexa group gives a route to combined DXP, CDP and customer engagement tooling. (+) AI generation and translation included in every plan.
Considerations: (-) Agentic AI features are in beta on the Ibexa platform; confirm general availability dates. (-) The Standard plan caps catalogues at 50,000 products and is seat-based. (-) GS1 GDSN and BMEcat support is not documented; manufacturers syndicating to retailers or German B2B channels should verify. (-) Integration into the Ibexa group may change roadmap priorities and branding; ask about product continuity.
Best for: Mid-market consumer brands in France and wider Europe that want a no-code PIM and DAM with fast onboarding.

## Vendor: Sales Layer
URL: https://pimatlas.com/vendors/sales-layer/
Product: Sales Layer PIM | Category: Mid-market PIM | HQ: Valencia, Spain | Founded: 2013 | Ownership: Private, venture backed (investors listed: PeakSpan Capital, Sonae/Bright Pixel, Swanlaab, Plug and Play); no acquisition verified as of October 2026
Deployment: SaaS | Segments: Mid-market, SMB | Standards: BMEcat, GS1, UNSPSC
Pricing model: Quote-based, per seat, billed annually or monthly. Four plans: Scale (5 users, 10,000 SKUs), Premium (10 users, 50,000 SKUs), Enterprise (35 users, 200,000 SKUs) and Enterprise Plus (custom). AI enablement is an add-on on Scale and included from Premium.

Sales Layer is a Spanish SaaS PIM aimed at mid-market manufacturers, distributors and retailers, especially in B2B sectors such as building materials, electrical and industrial equipment. It distinguishes itself with fast self-service onboarding, a free trial, tiered per-seat plans, an agentic AI layer with human review, an MCP server and procurement features such as punchout and BMEcat export.

Sales Layer was founded in 2013 by Álvaro Verdoy and Iban Borràs and is headquartered in Valencia, Spain. It raised a USD 25 million Series B in June 2022 led by PeakSpan Capital, with Bright Pixel (Sonae) and Swanlaab participating, and opened a UK office the same year. The company remains independent; PIM Atlas found no verified acquisition as of October 2026.

The product is a cloud PIM with a product toolkit, built-in DAM, connectors, instant catalogues and PDF export. Connectors cover Shopify, Adobe Commerce, WooCommerce, PrestaShop and BigCommerce, plus feed tools such as ChannelAdvisor, Lengow and Feedonomics. For B2B and procurement, Sales Layer offers punchout, BMEcat export and UNSPSC classification, and a GS1 Connector developed with partner Qumica publishes validated data to GS1 data pools.

AI is positioned as agentic. Individual AI actions are chained into agents, with Review Mode and simulation previews to keep a human in the loop, and an MCP server can be hosted by Sales Layer or installed locally. Plans are per seat with SKU and language limits, and a 30-day free trial is available.

Strengths: (+) Self-service onboarding with a 30-day free trial, which shortens evaluation. (+) Agentic AI actions with Review Mode and simulation previews for human approval. (+) MCP server in read-only and read/write modes, hostable by the vendor or locally. (+) B2B procurement features, including punchout, BMEcat export and UNSPSC categorisation. (+) ISO 27001 certification and a public trust portal.
Considerations: (-) Plans cap SKUs (10,000 to 200,000) and users below Enterprise Plus, so growth can trigger plan changes. (-) ERP connectors are not named on the connectors page; ERP integration typically runs through the API or partners. (-) GS1 data pool publishing relies on a partner-built connector (Qumica); ETIM is not documented. (-) No public prices despite a published plan structure.
Best for: Mid-market B2B manufacturers and distributors that want a quick-to-deploy SaaS PIM with agentic AI and procurement channel support.

## Vendor: Salsify
URL: https://pimatlas.com/vendors/salsify/
Product: Salsify PXM Platform | Category: PXM and syndication | HQ: Boston, United States | Founded: 2012 | Ownership: Private; Cinven agreed to acquire Salsify on 22 July 2026 (subject to regulatory approvals, completion not announced at last verification). Earlier investors include TPG and Warburg Pincus.
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1/GDSN, OpenAI Agentic Commerce Protocol
Pricing model: Quote-based subscription; no public list prices were verified for this profile.

Salsify is a Boston-based PXM (product experience management) platform that combines PIM, syndication to retailer and marketplace destinations, a GS1 GDSN data pool, enhanced content and supplier onboarding. It suits consumer brands and manufacturers that sell through many retail channels, especially in CPG and grocery. Salsify reports more than 2,000 customers and about 2,600 commerce destinations. In July 2026 the private equity firm Cinven agreed to acquire the company, subject to regulatory approvals.

Salsify was founded in 2012 in Boston and positions itself as a product experience management vendor for brand manufacturers, distributors and retailers. Its platform covers PIM, syndication, a GDSN data pool, enhanced content, digital shelf analytics, catalogue sites and, through the SXM product line, supplier onboarding for retailers. The 2021 acquisition of Paris-based Alkemics strengthened its supplier and retailer network in France and the rest of Europe.

Scale is central to the proposition. In its July 2026 deal announcement Salsify cited more than 2,000 customers, about 750 million products published to 2,600 commerce destinations and more than 70,000 active users. AI features sit inside the workflow engine: the Intelligence Suite (October 2025), the Angie assistant, FeedbackIQ for GDSN errors and, from May 2026, SalsifyIQ with an MCP layer for external AI agents. An OpenAI channel publishes brand-approved data using the Agentic Commerce Protocol.

Ownership is changing. On 22 July 2026 Cinven agreed to acquire Salsify for undisclosed terms; the transaction was subject to regulatory approvals, and no completion had been announced on Salsify's press page at the time of verification. Buyers signing multi-year contracts should ask about roadmap and pricing commitments under the new owner. The September 2026 IDEA Connector integration extends Salsify into North American electrical distribution.

Strengths: (+) Large syndication network: about 2,600 commerce destinations and 750 million published products (July 2026). (+) Native GDSN data pool, extended in 2025 to seven additional European markets. (+) AI embedded in workflows: Intelligence Suite, Angie, FeedbackIQ and the SalsifyIQ knowledge layer with MCP access. (+) Strong CPG and grocery references, including Mars, L'Oréal, The Coca-Cola Company, Carrefour and E.Leclerc. (+) Supplier experience (SXM) products for retailers that onboard supplier content at scale.
Considerations: (-) Pending change of control to Cinven (announced 22 July 2026); confirm contract protections and roadmap commitments. (-) Pricing is quote-based; syndication destinations, GDSN, enhanced content and AI modules can be packaged separately. (-) Support for B2B classification standards such as ETIM or ECLASS was not confirmed for this profile; manufacturers with technical catalogues should test it. (-) Hosting regions and EU data residency options were not verified; check them during due diligence.
Best for: Consumer brands and manufacturers that need to syndicate product content and GDSN data to many retailers and marketplaces from one platform.

## Vendor: SAP
URL: https://pimatlas.com/vendors/sap-master-data-governance/
Product: SAP Master Data Governance | Category: Product MDM | HQ: Walldorf, Germany | Founded: 1972 | Ownership: Public, listed on the Frankfurt Stock Exchange and NYSE (ADR)
Deployment: On-premise, Private cloud, SaaS | Segments: Enterprise | Standards: None verified
Pricing model: Quote-based; licensed through SAP contracts.

SAP Master Data Governance is SAP's master data management application for governing product, customer, supplier, business partner and finance master data. It suits large SAP-centric enterprises that need to create, approve and distribute product master records consistently across ERP landscapes. It runs on SAP S/4HANA or SAP Business Technology Platform and is typically paired with a PIM that manages marketing-facing product content.

SAP Master Data Governance (MDG) is a multidomain master data management application from SAP SE, the Walldorf-based enterprise software company. SAP positions it for a harmonised master data strategy across domains, with the product, customer, supplier, business partner and finance data typical of ERP-led organisations. It is available on SAP S/4HANA and on SAP Business Technology Platform, with on-premise, private cloud and public cloud deployment and support for hybrid landscapes that include non-SAP systems.

For product data, MDG governs the creation and change of material and product master records through central workflows, consolidation of records from multiple sources and mass processing. This makes it the system of record for operational product attributes used in procurement, manufacturing, logistics and finance. Marketing descriptions, rich media and channel-specific content are usually managed in a dedicated PIM fed by MDG.

SAP announced in June 2025 that MDG was named a Leader in The Forrester Wave: Master Data Management Solutions, Q2 2025. European PIM buyers in SAP landscapes often meet MDG during scoping, so the boundary between MDG and the PIM, and which system owns each attribute, should be agreed early.

Strengths: (+) Native governance of product master data inside SAP S/4HANA processes (+) Multidomain coverage including product, customer, supplier, business partner and finance data (+) On-premise, private cloud and public cloud deployment options, including SAP Business Technology Platform (+) Named a Leader in The Forrester Wave: Master Data Management Solutions, Q2 2025 (+) Large European partner ecosystem for implementation and support
Considerations: (-) Designed for operational master data, so marketing content, rich media and syndication need a separate PIM (-) Value depends heavily on an SAP-centric landscape and SAP skills (-) Shipped AI features specific to MDG were not verifiable from public product pages (-) Pricing is not published and licensing follows SAP contract structures
Best for: Large manufacturers and consumer goods companies running SAP S/4HANA that need governed product master data across ERP processes, alongside a PIM for channel content.

## Vendor: Semarchy
URL: https://pimatlas.com/vendors/semarchy/
Product: Semarchy Data Platform | Category: Product MDM | HQ: Phoenix, United States | Founded: 2011 (Lyon, France) | Ownership: Private; growth investment from funds advised by Providence Strategic Growth (2020), stake not disclosed
Deployment: SaaS, Private cloud, On-premise | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based; no prices are published. A free trial is offered, and the platform is listed on the Snowflake, Microsoft Azure, AWS and Google Cloud marketplaces.

Semarchy Data Platform (SDP) is a multi-domain master data management (MDM) and data governance platform from Semarchy, founded in Lyon in 2011 and now headquartered in Phoenix, Arizona. It suits enterprises that want product, customer and supplier data mastered on one platform with close ties to Snowflake and Microsoft Fabric. Gartner named Semarchy a Leader in its 2026 Magic Quadrant for Master Data Management, and SDP runs as SaaS, self-hosted or as a Snowflake Native App.

Semarchy launched the Semarchy Data Platform (SDP) in 2025; it previously marketed its platform as Semarchy xDM and the Intelligent Data Hub, and acquired the Lyon data integration vendor Stambia in 2021. SDP supports registry, consolidation, coexistence and centralised MDM styles in one model and adds data quality, governance and DataOps tooling for building governed data products. Product data is one domain among several, supported by pre-built product, customer and vendor models in the Accelerator Model Marketplace launched in February 2026.

Deployment choice is central to the offer: SaaS, a self-hosted option for on-premise or customer-managed cloud (including Red Hat OpenShift and SUSE Rancher RKE2), and a native application on Snowflake. Semarchy added a Fabric CLI in March 2026 and in September 2026 announced a Data Products workload for Microsoft Fabric that publishes mastered data to OneLake and lineage to Microsoft Purview.

Gartner named Semarchy a Leader in the 2026 Magic Quadrant for Master Data Management (April 2026). Funds advised by Providence Strategic Growth made a growth investment in 2020. Semarchy does not position SDP as a PIM, so channel syndication, digital asset management and marketing enrichment usually require a separate PIM or syndication tool.

Strengths: (+) Supports all four MDM implementation styles in a single model and codebase (+) Wide deployment choice: SaaS, self-hosted on-premise or in customer cloud, and a Snowflake Native App (+) Microsoft Fabric, OneLake and Purview integration, with a Fabric workload announced in September 2026 (+) Named a Leader in the 2026 Gartner Magic Quadrant for Master Data Management (+) Accelerator Model Marketplace with pre-built product, customer and vendor models that customers can extend
Considerations: (-) No PIM-style channel syndication or DAM is documented; plan a separate PIM or syndication layer for commerce channels (-) Prices are not published; costs depend on deployment model and scope (-) No product data standards such as GS1/GDSN or ETIM are documented on the Semarchy website (-) Several AI and Fabric capabilities were announced during 2026; check general availability for each
Best for: Enterprises that need multi-domain master data, including product, governed on Snowflake or Microsoft Fabric with a choice of SaaS or self-hosted deployment.

## Vendor: Stibo Systems
URL: https://pimatlas.com/vendors/stibo-systems/
Product: STEP (Product Experience Data Cloud) | Category: Product MDM | HQ: Aarhus, Denmark | Founded: 1976 | Ownership: Foundation-owned; subsidiary of Stibo Software Group within the Danish Stibo Group
Deployment: SaaS, On-premise | Segments: Enterprise | Standards: PIES 8.0
Pricing model: Quote-based; no public prices. ProductGen AI is a separate licence and is required for CoPilot and AI Prompt Creation features.

Stibo Systems is a Danish, foundation-owned master data management vendor whose STEP platform covers product, customer, supplier, location and sustainability data, with the Product Experience Data Cloud as its PIM offer. It suits large enterprises that want product information governed inside a multidomain MDM. Gartner named Stibo Systems a Leader in both its 2026 MDM and 2026 PIM Magic Quadrants, and fiscal 2025/26 revenue reached DKK 1.31 billion.

Stibo Systems was founded in 1976 as part of the Stibo Group, a Danish business that began as a printing house in 1794, and is headquartered in Aarhus. It is a subsidiary of Stibo Software Group and is foundation-owned, which the company presents as a basis for long-term independence. The STEP platform manages several master data domains; product data is sold as the Product Experience Data Cloud, alongside customer, supplier, location and sustainability data clouds.

The company is growing steadily from a large base. For fiscal year 2025/26 it reported revenue of DKK 1.31 billion (up 6.0%), SaaS revenue growth of 17.0%, an EBITDA margin of 18.1% and 77% customer adoption of its SaaS platform. Releases ship quarterly, with STEP 2026.3 released on 23 September 2026, and several 2025 releases were SaaS-only, so new capabilities reach cloud customers first.

AI work concentrates on governed use of master data. ProductGen AI generates and translates product content with approval workflows, the MCP Server became generally available in release 2026.2, and AgentWorkx, announced in September 2026, is a framework for building and governing agents inside STEP. Buyers should expect an enterprise MDM programme delivered by Stibo Systems and partners such as Nexer, with quote-based pricing.

Strengths: (+) Leader in the 2026 Gartner Magic Quadrants for both Master Data Management Solutions and Product Information Management Solutions. (+) Multidomain scope: product, customer, supplier, location and sustainability data on one STEP platform. (+) Financial stability: foundation ownership and DKK 1.31 billion revenue with an 18.1% EBITDA margin in fiscal 2025/26. (+) Governed AI stack: ProductGen AI, a generally available MCP Server and the AgentWorkx agent framework. (+) Quarterly release cadence with published release dates through 2027.
Considerations: (-) An enterprise MDM platform; scope, implementation effort and cost exceed the needs of many mid-market PIM projects. (-) ProductGen AI is a separately licensed add-on, and AgentWorkx agents were at early-interest stage in September 2026. (-) Several recent releases were SaaS-only; on-premise customers should check feature parity and upgrade paths. (-) No public pricing; total cost depends on domains, data volumes and modules.
Best for: Large enterprises that want product information managed and governed within a multidomain MDM alongside customer, supplier and location data.

## Vendor: Struct
URL: https://pimatlas.com/vendors/struct/
Product: Struct PIM | Category: Mid-market PIM | HQ: Odense, Denmark | Founded: Not disclosed | Ownership: Private, majority owned by Danish private equity firm VIA equity (about 75%) since 2022; no later ownership change verified
Deployment: SaaS | Segments: Mid-market, SMB | Standards: ETIM
Pricing model: Published monthly tiers with unlimited SKUs and assets: Basic from EUR 2,200, Professional from EUR 3,510 and Premium from EUR 4,800, with Enterprise quote-based. Add-ons such as the Supplier Onboarding Portal (EUR 350), ETIM Connector App (EUR 230) and Shopify Connector App (EUR 340) are priced per month.

Struct PIM is a Danish SaaS product information management system for mid-sized retailers, wholesalers and manufacturers, with a strong home market in Denmark and growing Nordic ambitions. It distinguishes itself through transparent published pricing with unlimited SKUs, a broad connector catalogue of more than 50 integrations and a practical AI toolkit that includes data extraction, image editing and an MCP server.

Struct A/S is based in Odense, Denmark, and describes itself as the market leader for PIM in Denmark. Danish private equity firm VIA equity acquired a majority stake of about 75% in 2022. The company reported its strongest year to date for 2025, with profit of DKK 5.45 million, more than double the prior year, and stated an ambition to become the preferred PIM platform in the Nordics before expanding further into Europe.

Struct PIM 4, announced in November 2024, moved the product to a multi-tenant cloud architecture on Microsoft Azure with a redesigned interface. The platform covers core PIM, a media library, business rules, workflows, publications for print, a Supplier Onboarding Portal and an Asset Sharing Portal. More than 50 connectors span e-commerce (Shopify, Adobe Commerce, commercetools, BigCommerce), ERP (SAP Business One, Business Central, Infor M3, IFS), DAM (Bynder, QBank) and marketplaces.

AI capabilities have been added steadily: the AI Assistant, Data Extraction, AI Image Manipulation and, in September 2026, an MCP server. Pricing is published in euros, which is uncommon in this market and simplifies early budgeting.

Strengths: (+) Transparent published pricing with unlimited SKUs, catalogues, assets and markets on every tier. (+) Broad connector catalogue of more than 50 integrations across e-commerce, ERP, DAM, CMS and marketplaces. (+) Practical shipped AI features: AI Assistant, Data Extraction, AI Image Manipulation and an MCP server. (+) Multi-tenant Azure architecture since Struct PIM 4, with two environments and continuous backup included. (+) Financially profitable vendor with a strong Danish customer base including Royal Unibrew, Kompan and Bolia.
Considerations: (-) Presence outside Denmark and Sweden is limited; buyers elsewhere should check local partner coverage and language support. (-) GS1 and BMEcat are not listed among connectors; ETIM is available as a paid connector app. (-) Users and languages are capped on Basic and Professional tiers, with extra users at EUR 70 and languages at EUR 200 per month. (-) Founding year and headcount are not published by the vendor.
Best for: Nordic mid-market retailers, wholesalers and manufacturers that want a quick-to-adopt SaaS PIM with predictable pricing.

## Vendor: SyncForce
URL: https://pimatlas.com/vendors/syncforce/
Product: SyncForce Circular PIM | Category: Mid-market PIM | HQ: Eindhoven, Netherlands | Founded: 1999 | Ownership: Private (owners not disclosed)
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1/GDSN, GS1 Digital Link, ETIM
Pricing model: Quote-based. Three licence plans (Essential, Professional, Ultimate) with unlimited users, products and digital assets; pricing tailored to company size and revenue.

SyncForce Circular PIM is a Dutch product information platform built specifically for packaged goods manufacturers, combining PIM, packaging data, DAM and stage-gate innovation management. It suits food, consumer goods and industrial brand owners that syndicate to retailers through GS1 GDSN. Distinguishing traits are its product, sales unit and transport unit data model, unlimited-user licences and modules for PPWR and EU Digital Product Passport readiness.

SyncForce traces its origins to 1999, when founder Hans de Gier built a cloud-based archive-sharing service for three brand manufacturers. The company is headquartered in Eindhoven, the Netherlands, and was named a Gartner Cool Vendor for Consumer Goods Manufacturers in 2014. Customer logos on its site include McCain Foods, Tyson Foods, Agristo, Brabantia, Bolsius and Dunlop Boots.

The product, branded Circular PIM, models data around a Golden Hierarchy that separates the core product from sales units and transport units, and a Triple Life Cycle covering supply, trade per market and data per channel. A Data Distribution Center transforms and validates output per channel, including a GS1 GDSN connector and requirements for GS1 Netherlands and GS1 Germany. Adaptive Change Management adds stage-gate workflows for new product development and tenders.

Regulatory modules address packaging and sustainability: the Ultimate plan includes a PPWR compliance manager, and an EU-DPP module prepares machine-readable datasets for ESPR. QR codes are managed centrally using GS1 Digital Link. Licences come in three plans with unlimited users and products, priced on company size and revenue.

Strengths: (+) Data model designed for packaged goods, separating product, sales unit and transport unit levels. (+) Native GS1 GDSN connector and GS1 Digital Link QR code management. (+) Packaging and regulatory modules, including a PPWR compliance manager and EU-DPP datasets. (+) Licences include unlimited users, products and digital assets. (+) Stage-gate innovation and change management in the same platform as product data.
Considerations: (-) Shipped AI features are not documented on the vendor's site; buyers prioritising AI enrichment should ask for a roadmap. (-) The specialist focus on packaged goods makes it a weaker fit for retailers or fashion brands. (-) Named e-commerce and ERP connectors are not published; integration relies on APIs and data ports. (-) No public prices; pricing is tied to company size and revenue.
Best for: Food, consumer goods and industrial brand manufacturers that syndicate to retailers via GS1 GDSN and must prepare packaging data for EU regulation.

## Vendor: Syndigo
URL: https://pimatlas.com/vendors/syndigo/
Product: Syndigo Product Experience Cloud (Synapse) | Category: PXM and syndication | HQ: Chicago, United States | Founded: Not disclosed | Ownership: Private equity backed by Summit Partners, TJC and Battery Ventures; acquired 1WorldSync (September 2025) and Taggstar (March 2026)
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: GS1/GDSN, OpenAI Agentic Commerce Protocol
Pricing model: Quote-based subscription; no public list prices were verified for this profile.

Syndigo is a Chicago-based product experience management vendor that combines PIM, multidomain MDM, syndication and, since acquiring 1WorldSync in September 2025, one of the largest GDSN commerce data pools. It connects more than 15,000 brands with over 3,500 retailers, distributors and platforms. It suits CPG, food and retail organisations that need content exchange between suppliers and retailers. Gartner named Syndigo a Leader in its inaugural 2026 Magic Quadrant for Product Information Management Solutions.

Syndigo is a private equity backed software company headquartered in Chicago and owned by Summit Partners, TJC and Battery Ventures. On 3 September 2025 it announced the acquisition of 1WorldSync, a commerce data pool and content network previously backed by Battery Ventures. The combined business was valued at more than US$3.5 billion and reported over 1,500 employees, 12 offices and more than 18,000 customers in 60 countries. In March 2026 Syndigo added Taggstar, an AI-driven conversion messaging company.

The Product Experience Cloud, sold in editions for brands and for retail, covers PIM, multidomain MDM, syndication, a GDSN data pool, rich media, PowerReviews ratings and reviews, analytics and Taggstar messaging. The commerce network is the main differentiator: brands publish once to a large set of retailer recipients, and retailers onboard supplier content through the same network.

AI is packaged around Synapse, launched in March 2026 as an agentic foundation on the MDM layer, and SynapseGo, which added conversational control and prebuilt agents in April 2026. Agentic Commerce Solutions include OpenAI Connect for the Agentic Commerce Protocol and generative engine optimisation services. Buyers should check how quickly 1WorldSync and Syndigo customers are moved to a single platform and how GDSN services are packaged.

Strengths: (+) Named a Leader in the inaugural Gartner Magic Quadrant for Product Information Management Solutions (report dated 12 October 2026). (+) Large brand and retailer network: more than 15,000 brands and 3,500 retailers, distributors and platforms. (+) GDSN data pool capacity strengthened by the 1WorldSync acquisition. (+) Breadth beyond PIM: multidomain MDM, rich media, PowerReviews ratings and reviews and Taggstar conversion messaging. (+) Agentic AI with human approval through Synapse and SynapseGo prebuilt agents.
Considerations: (-) Post-merger integration of 1WorldSync is recent; confirm platform roadmap, migration plans and support model for legacy 1WorldSync products. (-) The portfolio is broad and modular; scope and licensing can be complex to compare against single-product PIM vendors. (-) European office footprint and EU hosting options were not verified; check data residency during due diligence. (-) Support for B2B standards such as ETIM or ECLASS was not confirmed for this profile.
Best for: CPG, food and retail organisations that exchange product content with many trading partners and need PIM, GDSN and syndication on one network.

## Vendor: THRON
URL: https://pimatlas.com/vendors/thron/
Product: THRON Platform (Digital Asset Library and Product Information Library) | Category: DAM and PIM suite | HQ: Piazzola sul Brenta, Italy | Founded: Not disclosed | Ownership: Private, THRON S.p.A. (shareholders not disclosed)
Deployment: SaaS | Segments: Enterprise, Mid-market | Standards: None verified
Pricing model: Quote-based modular pricing: a Start package extended with features or bundles (Marketing, Sales, Corporate); several AMBRA AI features are add-ons with Pro and Ultra tiers. Free trial available.

THRON is an Italian SaaS platform that combines DAM and PIM, with DAM as its origin and centre of gravity. It suits Italian and European brands in fashion, furniture, ceramics and appliances that manage large volumes of product imagery alongside product sheets. Its distinguishing traits are the AMBRA AI suite for asset enrichment, image editing and product descriptions, built-in CDN delivery and an MCP interface for external AI tools.

THRON S.p.A. is headquartered in Piazzola sul Brenta near Padova, Italy. The company began as a digital agency around the turn of the millennium, and its founders include Nicola Meneghello (President) and Alberto Mosca. Customer logos include Whirlpool, Valentino, Barilla, Dainese, Geox, Iveco Group, Scavolini, Ariston and Atlas Concorde, which reflects a strong base among Italian manufacturers and fashion houses.

The platform is a geo-distributed SaaS built on microservices. Its Digital Asset Library handles asset cataloguing, rights, workflow and delivery through a CDN, while the Product Information Library manages structured product sheets. Add-ons include THRON Drive, Dynamic Print Publishing, a B2B Area and a Brand Portal. Named connectors cover Salesforce Commerce Cloud, Adobe Commerce, Shopify, Adobe Experience Manager, WordPress, Drupal and Adobe creative tools.

AMBRA AI, presented in April 2025, now spans search, enrichment, descriptions, translation, accessibility, image editing, data modelling and an MCP interface. The vendor positions the product as agentic SaaS. Pricing is modular and quote-based, with several AI functions sold as add-ons.

Strengths: (+) Combined DAM and PIM on one SaaS platform, with built-in CDN delivery of assets up to 4K. (+) Extensive named AI suite (AMBRA AI) covering enrichment, descriptions, translation, accessibility and image editing. (+) AI MCP interface that exposes content to external assistants such as ChatGPT, Claude and Gemini. (+) ISO 27001, ISO 27017 and ISO 27018 certification. (+) Reference base of Italian and European manufacturers and fashion brands.
Considerations: (-) PIM depth is lighter than at PIM-first vendors; buyers with complex B2B data models should test variant handling and classification. (-) No support for GS1, ETIM or BMEcat is documented. (-) Named ERP connectors are not published; ERP integration runs through the API framework. (-) Several AI functions are add-ons, so AI costs depend on the chosen tier.
Best for: European brands, especially Italian manufacturers and fashion houses, whose product content is image-heavy and centred on DAM.

## Vendor: Viamedici
URL: https://pimatlas.com/vendors/viamedici/
Product: Viamedici EPIM (EPIM/5) | Category: Enterprise PIM | HQ: Ettlingen, Germany | Founded: 1999 | Ownership: Private, independent (Viamedici Software GmbH; no external investors disclosed)
Deployment: SaaS, Private cloud, On-premise | Segments: Enterprise, Mid-market | Standards: ETIM, ECLASS, GS1
Pricing model: Quote-based; modules (VIA/MDM, VIA/PIM360°, VIA/DAM, VIA/Configuration) can be licensed separately and combined. No public prices.

Viamedici EPIM is a German multidomain PIM and MDM platform for manufacturers with complex, variant-rich product ranges. EPIM/5 combines PIM, master data management, DAM and product configuration (CPQ) on one data model, runs on AWS, Azure, a vendor-operated sovereign cloud or on-premise, and is SAP-certified. Gartner named Viamedici a Visionary in its 2026 Magic Quadrant for Product Information Management Solutions, making it a credible option for industrial manufacturers with SAP landscapes.

Viamedici Software GmbH was founded in Germany in 1999 and is headquartered in Ettlingen, near Karlsruhe. The company describes itself as privately held and reports more than 400 customers. Its EPIM platform, currently in its fifth generation (EPIM/5), is sold as a set of modules: VIA/MDM for multidomain master data, VIA/PIM360° for product information, VIA/DAM for digital assets and VIA/Configuration for guided selling and CPQ. Buyers can start with one module and add others later.

The product targets mid-sized companies and international groups with large data volumes, many users and multilingual requirements. Named customers include TRUMPF, SICK, Baumit, ZARGES and Reichle & De-Massari, which reflects a focus on industrial manufacturing, building products and technical goods. The platform supports ETIM, ECLASS and GS1 classification, includes a supplier portal and print output for catalogues, and offers bidirectional SAP integration.

In 2026 Viamedici received two analyst placements: Niche Player in the Gartner Magic Quadrant for Master Data Management Solutions (April 2026) and Visionary in the Gartner Magic Quadrant for Product Information Management Solutions (October 2026). EPIM/5 Update 5, announced on 1 October 2026, added an Event Hub that pushes data changes to connected systems in real time, the Data Expert AI chat assistant and an AI translation service. Sustain360 adds digital product passport generation, relevant to buyers preparing for the Ecodesign for Sustainable Products Regulation (Regulation (EU) 2024/1781).

Strengths: (+) Multidomain data model covering product, supplier, customer and location master data alongside PIM, DAM and CPQ on one platform (+) SAP-certified with bidirectional SAP integration, suited to SAP-centred manufacturers (+) Flexible hosting on AWS, Azure, a vendor-operated sovereign cloud, hybrid or on-premise (+) Named a Visionary in the 2026 Gartner Magic Quadrant for PIM Solutions and a Niche Player in the 2026 Gartner Magic Quadrant for MDM Solutions (+) Shipped AI features in EPIM/5 Update 5, including Data Expert AI and an AI translation service with CLDR locale support
Considerations: (-) Breadth of modules implies a substantial implementation; buyers needing only basic PIM should assess scope and cost (-) No public pricing; licence cost depends on module mix and hosting choice (-) BMEcat and GS1 GDSN support are not documented on the vendor website and should be confirmed in an RFP (-) Most named customers are DACH industrial firms, so references in retail or fashion should be requested
Best for: Industrial and technical manufacturers with SAP landscapes that want PIM, multidomain MDM, DAM and product configuration from one vendor.

## Vendor: W4
URL: https://pimatlas.com/vendors/elytra-pim/
Product: Elytra PIM (formerly Lobster_pim) | Category: Mid-market PIM | HQ: Zurich, Switzerland | Founded: 1994 (W4); the product was originally developed by Lobster DATA GmbH | Ownership: Private, family-owned (W4, founded by Ivana and Jörg Wenzel)
Deployment: SaaS, Private cloud, On-premise | Segments: Mid-market, Enterprise | Standards: BMEcat
Pricing model: Quote-based. No public prices; existing Lobster_pim contracts continue unchanged under W4.

Elytra PIM is the continuation of Lobster_pim, whose full source code Swiss company W4 acquired from Lobster DATA GmbH in June 2025 and renamed in Release 25.0. It is a no-code, configurable PIM for mid-sized and large companies, available on-premise, in a private cloud or as SaaS. Customers inherited from Lobster include Lidl, Beckhoff and Galenica. Elytra PIM suits existing Lobster_pim users and DACH buyers who prioritise sovereign hosting.

Lobster_pim was developed by Lobster DATA GmbH, the German data integration company based in Tutzing. In June 2025 W4, a family-owned digital agency and software company with roots in Zurich, acquired the complete source code with all components and functionality. Lobster refocused on data integration and process automation, and its current website lists only the Lobster Data Platform. W4 had been part of Lobster's partner network for many years through its brand wps4digital.

W4 renamed the product Elytra PIM in Release 25.0 (Q3 2025), alongside web client bug fixes, and announced RESTful product data APIs based on OpenAPI for Release 25.1 (Q4 2025). Existing customer contracts concluded with Lobster continue unchanged, and W4 states that integrations with Lobster_data, priint:comet, Canto and SixOMC are maintained. A dedicated support and development team works in German and English.

Functionally, the product offers central product data management with a no-code configurable data model, multilingual content, validation rules, role-based workflows and automated publishing to print, e-commerce and other channels. Legacy Lobster_pim documentation lists BMEcat output and priint:comet print integration. Named customers include Lidl, Beckhoff, Galenica, Fenaco, Böllhoff, Bystronic and Thonet, and W4 cites typical implementations of four to eight months. Hosting is offered on-premise, in a private cloud or as SaaS, with an emphasis on sovereign hosting.

Strengths: (+) Established product with an installed base inherited from Lobster, including Lidl, Beckhoff and Galenica (+) Choice of on-premise, private cloud or SaaS, with sovereign hosting positioned for DACH buyers (+) No-code configuration of the data model, workflows and validation rules (+) Continuity for existing customers: Lobster contracts and key integrations (Lobster_data, priint:comet, Canto) are maintained (+) Typical implementation time of four to eight months according to W4
Considerations: (-) Product ownership changed in 2025; buyers should review W4's roadmap, release cadence and development capacity (-) No shipped AI features are documented (-) Standards support beyond BMEcat (ETIM, ECLASS, GS1) is not documented and should be confirmed (-) Delivery of the OpenAPI-based REST API announced for Release 25.1 should be verified
Best for: Existing Lobster_pim customers and DACH mid-sized companies that want a configurable PIM with on-premise or sovereign hosting options.

## Vendor: Webkul
URL: https://pimatlas.com/vendors/unopim/
Product: UnoPim | Category: Open source PIM | HQ: Noida, India | Founded: 2024 (UnoPim launch); Webkul founding year not verified | Ownership: Private; open source project developed by Webkul
Deployment: On-premise, SaaS | Segments: Mid-market, SMB | Standards: None verified
Pricing model: Open source core free under the MIT licence. Managed cloud hosting from USD 50 per month. Connectors (Shopify, Magento 2, WooCommerce, Odoo), DAM, PDF Generator and Maker Checker are paid extensions sold through the Webkul store; implementation is quote-based.

UnoPim is an MIT-licensed open source PIM and DAM platform built on Laravel by Webkul, the Indian e-commerce software developer. It suits small and mid-sized merchants and agencies that want a free, modern PHP codebase with strong AI tooling and low hosting cost. Its distinguishing features are the permissive MIT licence, multi-provider Magic AI and a natural-language AI agent built into the core.

UnoPim is an open source product information management system released by Webkul in 2024. It is built on Laravel 13 and PHP 8.4, runs on MySQL or PostgreSQL and is distributed under the MIT licence, which places fewer obligations on adopters than the copyleft GPL or OSL licences used by several other open source PIMs. The GitHub repository shows around 11,000 stars, and version 3.1.3 was published on 29 September 2026.

The core covers attributes, families, categories, multi-channel and multi-currency data, 33 interface languages, import and export jobs, webhooks and Digital Product Passport publishing. AI is a central theme of the 3.x line: Magic AI supports more than ten model providers, and an AI Agent Chat performs catalogue tasks such as bulk edits and image-based product creation from natural-language instructions.

The commercial model follows Webkul's extension approach. The core is free, managed cloud hosting starts at USD 50 per month, and connectors for Shopify, Magento 2, WooCommerce and Odoo, along with DAM and approval workflow features, are sold as paid extensions. Webkul lists offices in Noida (India), Wilmington (USA) and Poznań (Poland) and publishes ISO 27001 and SOC 2 Type 2 attestations on the UnoPim site.

Strengths: (+) Permissive MIT licence allows modification and redistribution without copyleft obligations (+) Modern Laravel 13 and PHP 8.4 stack familiar to a large developer community (+) Multi-provider Magic AI and a natural-language AI agent included in the open source core (+) Low entry cost with managed hosting from USD 50 per month (+) Digital Product Passport publishing included in the core feature set
Considerations: (-) Young product launched in 2024 with a short track record in large enterprise rollouts (-) DAM, approval workflows and channel connectors are paid extensions (-) No verified support for ETIM, BMEcat or GS1/GDSN exchange (-) Product management and support run largely from India, so check European partner coverage
Best for: Small and mid-sized merchants and digital agencies that want a free, MIT-licensed PIM on Laravel with built-in AI and inexpensive connectors to Shopify, Magento, WooCommerce or Odoo.

## Article: What is PIM? A practical definition of product information management
URL: https://pimatlas.com/knowledge/what-is-pim/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: A PIM (product information management) system is the central application where a business collects, enriches, governs and publishes the marketing and technical information that describes its products. It sits between source systems such as ERP and the channels that sell, so every website, marketplace, catalogue and retailer receives consistent, complete product data.

A PIM (product information management) system is the single application where a business collects, enriches, governs and publishes the information that describes its products. It takes basic item data from source systems, adds the marketing and technical detail that buyers need, checks it against quality rules and sends a channel-ready version to every website, marketplace, print catalogue and trading partner.

Gartner, which published a Market Guide for Product Information Management Solutions on 22 January 2025, frames the category around a practical problem: digital commerce requires organisations to create, maintain and publish product information to many downstream channels. A PIM is the tool built to do that at scale.

## What does a PIM system actually store?

A PIM holds the descriptive layer of product data. It typically manages five kinds of information:

- **Identifiers:** SKU, GTIN (Global Trade Item Number, the GS1 identifier that is usually 13 digits in Europe), manufacturer part numbers.
- **Attributes:** dimensions, materials, technical specifications, regulated values such as allergens or energy class.
- **Content:** product names, short and long descriptions, bullet points, SEO text, all in every language you sell in.
- **Structure:** categories, classification codes (for example ETIM or GS1 GPC), variants and product relationships such as accessories and spare parts.
- **Media references:** links to images, videos, manuals and certificates, which usually live in a DAM (digital asset management) system.

Prices, stock levels and order history normally stay in ERP and commerce systems. The PIM may display them for context, but it rarely owns them. The article on [PIM and ERP](https://pimatlas.com/knowledge/pim-vs-erp/) sets out that split in detail.

## How does a PIM work day to day?

Most PIM implementations follow the same flow, regardless of vendor.

1. **Import.** Core item records arrive from ERP or PLM; supplier data arrives through files, portals or GDSN (the GS1 Global Data Synchronisation Network, which shares data through certified data pools).
2. **Model.** Products are assigned to families or classes that define which attributes each product type needs.
3. **Enrich.** Teams add copy, translations, attributes and media links, often supported by AI-assisted generation and classification.
4. **Validate.** Completeness rules and workflows check that each product meets the requirements of each channel and locale before release.
5. **Publish.** The PIM exports or syndicates channel-specific views to web shops, marketplaces, retailers, print and B2B portals through APIs, feeds or standard formats such as BMEcat.

The value sits in steps 4 and 5. Without them, every channel team maintains its own copy of the truth, and errors multiply with each new market.

## When does a business need a PIM?

Spreadsheets and ERP fields work for a small range sold through one channel. The table below lists the thresholds where most organisations start to struggle without a dedicated system.

| Signal | Typical threshold | Why it matters |
|---|---|---|
| Range size | Several thousand SKUs or more | Manual maintenance cannot keep attributes consistent |
| Channels | Three or more outlets with different data requirements | Each marketplace and retailer has its own attribute model |
| Languages | Two or more locales | Translation status needs tracking per attribute |
| Time to market | New products wait weeks for content | Lost sales during the enrichment backlog |
| Regulation | Mandatory online product information | Rules such as the EU General Product Safety Regulation require specific data at the point of sale |
| Supplier data | Many suppliers sending different formats | Onboarding becomes a full-time manual task |

Two or more of these signals usually justify a structured evaluation. The [business case article](https://pimatlas.com/knowledge/pim-business-case/) shows how to quantify them.

## What a PIM is not

A PIM sits in a crowded data stack, and its boundaries are often confused. It is a different system from an ERP, which runs transactions. It overlaps with product MDM (master data management), which focuses on governed, cross-domain master records, as explained in [PIM vs MDM](https://pimatlas.com/knowledge/pim-vs-mdm/). It also differs from a DAM, which stores and manages the media files themselves.

Many vendors now combine several of these capabilities in one suite, and some market their products as PXM (product experience management) platforms. The labels matter less than the scope: buyers should define which data the PIM will own, and which it will only reference, before comparing products in the [vendor directory](https://pimatlas.com/vendors/).

## Key takeaways

- A PIM system centralises, enriches, validates and publishes the descriptive information about products for every sales channel.
- It owns identifiers, attributes, content, classification and media references; prices, stock and transactions usually stay in ERP.
- The core workflow is import, model, enrich, validate and publish, with validation and publishing delivering most of the value.
- Range size, channel count, languages, time to market, regulation and supplier volume are the practical signals that a PIM is needed.
- Define data ownership across ERP, PIM, MDM and DAM before shortlisting vendors.

Q: What does PIM stand for?
A: PIM stands for product information management. It describes both the discipline of managing the information that describes products and the software category built for it. A PIM system centralises attributes, descriptions, classifications, relationships and links to media, then applies workflows and quality rules before publishing that information to e-commerce sites, marketplaces, print catalogues and trading partners.
Q: Who uses a PIM system?
A: Product data, e-commerce and category teams use a PIM daily to enrich and approve product content. Marketing teams write copy and attach media, compliance teams maintain regulated attributes, and IT teams own the integrations. Suppliers can also contribute data through onboarding portals. Manufacturers, brands, wholesalers, distributors and retailers all run PIM systems, usually once their range or channel count outgrows spreadsheets.
Q: Is a PIM the same as an ERP?
A: No. An ERP (enterprise resource planning) system runs transactions such as orders, inventory, purchasing and finance, and stores the operational subset of product data those processes need. A PIM manages the richer descriptive content that customers and channels need, such as marketing copy, detailed specifications, translations and channel-specific attributes. Most businesses run both and integrate them.

## Article: PIM vs MDM: where product information ends and master data begins
URL: https://pimatlas.com/knowledge/pim-vs-mdm/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: PIM manages the rich, channel-facing product content that customers and retailers see. MDM governs the core master records, such as products, customers, suppliers and locations, that operational systems share. Product MDM and PIM overlap on product identity and attributes, so most enterprises use MDM for governed core data and PIM for enrichment and publishing.

PIM and MDM solve related problems for different audiences. A PIM (product information management) system manages the rich, channel-facing content that describes products to customers and trading partners. MDM (master data management) governs the core records that many operational systems share, across domains such as product, customer, supplier and location. Where they meet, in product MDM, the dividing line is purpose: MDM makes the record correct and unique, and PIM makes it sellable.

## How do the definitions differ?

Gartner defines MDM as "a technology-enabled business discipline in which business and IT work together to ensure the uniformity, accuracy, stewardship, governance, semantic consistency and accountability of the enterprise's official shared master data assets". It defines master data as "the least number of consistent and uniform sets of identifiers and extended attributes that describe the core entities of an enterprise".

That definition highlights the key contrast. MDM aims for the minimum consistent set of identifiers and attributes that every system can trust. PIM aims for the maximum useful set of descriptive content that every channel needs. One reduces data to a governed core, the other expands it for commerce.

## PIM and MDM compared side by side

| Dimension | PIM | MDM (including product MDM) |
|---|---|---|
| Primary goal | Complete, channel-ready product content | One trusted, de-duplicated master record |
| Data domains | Product, sometimes supplier and asset links | Product, customer, supplier, location, asset, employee |
| Typical attributes | Marketing copy, specifications, translations, channel attributes | Identifiers, classifications, hierarchies, cross-references |
| Main users | E-commerce, marketing, category and product data teams | Data stewards, data governance office, IT |
| Core functions | Enrichment, completeness scoring, workflow, syndication | Matching, merging, survivorship, hierarchy management |
| Downstream consumers | Web shops, marketplaces, retailers, print, B2B portals | ERP, CRM, supply chain, analytics, data warehouse |
| Implementation style | Business-led, iterative by channel | Programme-led, often tied to ERP consolidation |

Some platforms span both columns. Vendors such as [Stibo Systems](https://pimatlas.com/vendors/stibo-systems/) and [Informatica](https://pimatlas.com/vendors/informatica/) position multidomain MDM alongside product content capabilities, while many PIM vendors add governance and matching features. Treat the table as a description of functions to cover, independent of which product delivers them.

## When is MDM the right starting point?

MDM earns its cost when the core problem is inconsistent identity across systems. Typical triggers include:

- Several ERP instances after mergers, each with its own item numbering.
- Duplicate supplier or customer records that distort purchasing and credit decisions.
- Analytics that cannot reconcile sales by product because hierarchies differ by country.
- Regulatory or financial reporting that depends on a single authoritative product hierarchy.

In these cases, adding a PIM on top of inconsistent identifiers only spreads the inconsistency to more channels.

## When is PIM the right starting point?

PIM is the better first investment when identifiers are stable and the bottleneck is content. Signals include long time to market for new products, poor marketplace acceptance rates, translation backlogs and retailer data requests that teams answer by spreadsheet. The [What is PIM](https://pimatlas.com/knowledge/what-is-pim/) article lists practical thresholds.

A PIM also delivers visible results faster. Teams can model one product family, connect one channel and measure the effect on content completeness within months. That shorter feedback loop makes PIM a common first step for mid-market manufacturers and brands, with MDM added later if identity problems grow as the business expands into new entities or regions.

## How do PIM and MDM work together?

A common enterprise pattern separates responsibilities by attribute group:

1. MDM or ERP creates the product record, assigns identifiers such as the GTIN and maintains the governed hierarchy.
2. PIM receives that record, adds commercial content, translations and channel attributes, and manages approval workflows.
3. PIM publishes to channels and, where required, to GS1 data pools for GDSN (Global Data Synchronisation Network) exchange.
4. Selected enriched attributes flow back to MDM or ERP only where downstream processes need them.

The critical design task is an attribute ownership matrix that names one system of record per attribute. The [governance article](https://pimatlas.com/knowledge/product-data-governance/) and the [integration architecture article](https://pimatlas.com/knowledge/pim-integration-architecture/) explain how to build and maintain it.

## Key takeaways

- MDM governs a minimal, trusted set of identifiers and attributes across many data domains; PIM enriches and publishes rich product content for channels.
- Start with MDM when identity is inconsistent across systems; start with PIM when identifiers are stable and content is the bottleneck.
- Product MDM and PIM overlap, and several vendors cover both, so define the functions you need before choosing a product category.
- An attribute ownership matrix with one system of record per attribute prevents conflicts between MDM, ERP and PIM.

Q: Is PIM a type of MDM?
A: PIM is often described as a product-domain specialisation of master data management, and the two share techniques such as data models, validation rules and stewardship workflows. In practice the categories differ in focus: MDM governs core identifiers and shared attributes across many systems and domains, while PIM concentrates on enriching and publishing customer-facing product content to sales channels.
Q: Do I need both PIM and MDM?
A: Organisations that run several ERP instances, need customer and supplier mastering, or must reconcile product records across business units usually benefit from MDM. Organisations whose main challenge is channel content, translation and syndication usually start with PIM. Large enterprises often run both, with MDM supplying governed core product records and PIM adding commercial content on top.
Q: Can a product MDM platform replace a PIM?
A: Some multidomain MDM platforms include product enrichment, workflow and publishing features that cover common PIM needs, and some PIM platforms add MDM-style matching and governance. Whether one tool can cover both depends on your channel requirements, user volumes and integration landscape. Test both scopes with scripted scenarios during vendor demonstrations before deciding.

## Article: PIM vs DAM: how the two systems divide the work
URL: https://pimatlas.com/knowledge/pim-vs-dam/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: A PIM manages structured product data such as attributes, descriptions and classifications. A DAM (digital asset management) system stores, organises, transforms and controls the rights to media files such as images, videos and documents. The PIM links each product to its approved assets, and the DAM delivers the right file in the right format to each channel.

PIM and DAM divide the work along a simple line: PIM manages structured product data, and DAM manages the media files that show and explain the product. A PIM (product information management) system holds attributes, descriptions and classifications. A DAM (digital asset management) system stores, organises, converts and governs images, videos, 3D models and documents. Product-led businesses get the best results when the PIM links each product to its approved assets and the DAM supplies those assets in the formats each channel requires.

## What does each system do?

The two categories overlap at the point where a product meets its pictures. The functions on either side of that point are distinct.

Gartner Peer Insights lists the mandatory features for its DAM market as a searchable repository, digital rights controls, single and bulk ingestion, usage reporting, approval workflows and integrations with marketing tools. None of these concern product attributes. A PIM, by contrast, centres on data models, attribute completeness, translation and channel syndication.

| Function | PIM | DAM |
|---|---|---|
| Core object | Product record (SKU, variant, family) | Asset file (image, video, PDF, 3D) |
| Main data | Attributes, copy, classifications, relationships | Files, renditions, technical and descriptive metadata |
| Rights and licences | Rarely managed | Usage rights, expiry dates, territories |
| Format conversion | Limited (thumbnails, basic resizing) | Resizing, cropping, transcoding, CDN delivery |
| Workflows | Enrichment and approval of product data | Creative review, brand approval, versioning |
| Primary users | Product data, e-commerce, category teams | Marketing, creative, brand, studio teams |
| Output | Channel-specific product data feeds | Channel-specific asset renditions and links |

## Where does the overlap cause problems?

Most friction appears in three places.

1. **Duplicate storage.** Images uploaded directly into the PIM and separately into the DAM drift apart, and teams publish outdated versions.
2. **Rights exposure.** A licensed lifestyle image with a 12-month term stays live on a marketplace because the PIM has no record of the expiry date. Rights belong in the DAM, as covered in the [rights management article](https://pimatlas.com/knowledge/dam-rights-management/).
3. **Unclear asset roles.** Channels need to know which image is the main image, which is the detail shot and which shows dimensions. If neither system owns that role, every export needs manual correction.

Regulation adds pressure. Under the EU General Product Safety Regulation (Regulation (EU) 2023/988), which applies from 13 December 2024, online offers must include information identifying the product, including a picture, alongside warnings and safety information. That makes the link between the product record and its approved images a compliance matter.

## How should PIM and DAM share the work?

A clean division assigns each piece of information one owner:

- **DAM owns** the master file, all renditions, technical metadata, descriptive tags, rights and expiry.
- **PIM owns** the product-to-asset relationship, the asset role per product, the sort order and any channel-specific selection.
- **Both share** a stable asset identifier and the product identifier (SKU or GTIN) so links survive file updates.

Integration then becomes a matter of synchronising references, with no file copying. The DAM publishes a URL or rendition reference, and the PIM attaches it to the product. The [DAM and PIM integration article](https://pimatlas.com/knowledge/dam-pim-integration/) describes the common patterns and their trade-offs.

Agree the naming convention early. A file name that encodes the SKU, the asset role and a sequence number lets both systems match assets automatically during bulk loads and photo studio deliveries.

## Which should come first?

Choose based on the bottleneck. If products wait for attributes, copy and translations, start with PIM and use its built-in asset features for packshots. If teams cannot find approved images, recreate assets or face licence breaches, start with DAM.

Volume is a useful test. A range with a few images per SKU and no video is often well served by PIM asset features. A library with tens of thousands of assets, multiple photo studios, video and licensed content usually needs a dedicated DAM. Clear [product image standards](https://pimatlas.com/knowledge/product-image-standards/) help in both cases, because they define what each system must deliver.

## Key takeaways

- PIM manages structured product data; DAM manages media files, renditions and usage rights.
- The overlap sits at the product-to-asset relationship, which should be owned by one system and referenced by the other.
- Rights, expiry dates and technical metadata belong in the DAM; asset roles and channel selection usually belong in the PIM.
- The EU GPSR requires product pictures and safety information in online offers, which makes reliable PIM and DAM links a compliance issue.
- Start with the system that removes your larger bottleneck, then integrate through shared identifiers.

Q: Can a PIM replace a DAM?
A: Many PIM systems include basic asset management: uploading images, linking them to products and generating channel renditions. That is often enough for a modest volume of packshots. Organisations with large libraries, video, creative workflows, brand guidelines or licensed imagery with expiry dates usually need a dedicated DAM, because those functions sit outside the core design of most PIM platforms.
Q: Should I buy PIM or DAM first?
A: Start with the system that addresses the larger bottleneck. If products wait for attributes, copy and translations, start with PIM. If teams cannot find approved images, recreate assets or misuse licensed content, start with DAM. Many product-led businesses implement PIM first, using its built-in asset features, and add a DAM when media volume or rights complexity grows.
Q: Where should product image metadata live?
A: Technical and rights metadata, such as file format, resolution, photographer, licence terms and expiry date, belongs in the DAM. The product relationship, such as which SKU an image shows, its role (main image, lifestyle, detail) and its sort order per channel, is usually managed in the PIM or synchronised between both systems through an agreed identifier.

## Article: PIM and ERP: which system should own which product data
URL: https://pimatlas.com/knowledge/pim-vs-erp/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: ERP should own the operational product data that transactions depend on: item numbers, units of measure, costs, prices, stock and logistics data. PIM should own the descriptive and commercial content that channels need: marketing copy, detailed specifications, translations, classifications and media links. One system of record per attribute, with ERP feeding PIM, prevents conflicts.

ERP should own the product data that transactions depend on, and PIM should own the product data that customers and channels depend on. In practice that means item numbers, units, costs, prices, stock and logistics live in ERP, while marketing copy, detailed specifications, translations, classifications and media links live in PIM. Each attribute gets one system of record, and ERP usually feeds the PIM.

## Why can't ERP handle product content on its own?

Gartner defines ERP (enterprise resource planning) as "the ability to deliver an integrated suite of business applications" that share "a common process and data model" across finance, HR, distribution, manufacturing, service and the supply chain. That data model is built for orders, invoices and stock movements. It is optimised for accuracy and control of transactions.

Product content has different needs. A power tool sold through a web shop, two marketplaces and a dealer portal may need 60 attributes on one channel and 120 on another, descriptions in eight languages, and different image sets per retailer. ERP item masters rarely offer flexible attribute sets per product type, rich text, translation workflows or completeness scoring by channel. Customising ERP to add them is possible, but changes to ERP are usually slow and expensive because every modification affects core business processes.

A PIM (product information management) system is built for that variability. The [What is PIM](https://pimatlas.com/knowledge/what-is-pim/) article describes its core functions.

## Who should own which attributes?

The table below shows a typical split. Exact boundaries vary by industry, so treat it as a starting template for your own ownership matrix.

| Attribute group | Typical system of record | Reason |
|---|---|---|
| Item number, SKU, status | ERP | Required for purchasing, sales orders and invoicing |
| GTIN (barcode number) | ERP or MDM | Assigned once and used in logistics and point of sale |
| Units of measure, pack sizes, weights for shipping | ERP | Drive warehouse, transport and customs processes |
| Cost, transactional price, discounts | ERP or pricing system | Determine margin and invoicing |
| Stock and availability | ERP or warehouse system | Change continuously with transactions |
| Marketing names, descriptions, bullet points | PIM | Written and approved by commercial teams |
| Detailed technical specifications | PIM (sometimes PLM) | Extensive, product-type specific, channel facing |
| Translations and localised content | PIM | Managed through translation workflows |
| Channel classifications (ETIM, GS1 GPC, marketplace categories) | PIM | Differ per channel and change often |
| Product relationships, accessories, spare parts | PIM | Commercial logic for cross-sell and service |
| Image and document links | PIM with DAM | Assets stored in DAM, roles assigned in PIM |

## How should data flow between ERP and PIM?

Most implementations use a one-way core flow with tightly controlled exceptions.

1. **ERP creates the item** with identifiers, units and logistics data once the product is approved for purchase.
2. **PIM receives the item** through an API or scheduled interface and assigns it to a product family.
3. **Teams enrich in PIM** and the PIM validates completeness per channel.
4. **PIM publishes** to channels, combining its content with price and stock references from ERP at publication time.
5. **Selected attributes return to ERP** only where a process needs them, such as a short description on invoices or customs codes.

Two-way synchronisation of the same attribute creates conflicts, so avoid it. The [integration architecture article](https://pimatlas.com/knowledge/pim-integration-architecture/) covers interface patterns, and the [governance article](https://pimatlas.com/knowledge/product-data-governance/) covers who approves changes to the matrix.

## What are the common mistakes?

Four errors recur in PIM and ERP projects:

- **Copying every ERP field into PIM.** Import only what enrichment or publishing needs; extra fields create noise and confusion about ownership.
- **Editing ERP data in PIM.** If users can change a weight in PIM, the warehouse and the web shop will soon disagree.
- **Ignoring product lifecycle status.** PIM must respect ERP status codes so discontinued items leave channels on time.
- **Leaving the matrix undocumented.** An ownership decision that lives only in an integration specification will be lost when the team changes.

## Key takeaways

- ERP owns transactional product data: identifiers, units, logistics, costs, prices and stock.
- PIM owns commercial and descriptive content: copy, specifications, translations, classifications, relationships and asset links.
- Document one system of record per attribute in an ownership matrix and review it when channels or systems change.
- Use a one-way core flow from ERP to PIM and return only the few attributes ERP processes require.
- Respect ERP lifecycle status in PIM so products enter and leave channels at the right time.

Q: Can ERP be used as a PIM?
A: ERP systems can store extra product fields, and some include basic catalogue functions. They are designed for transactions, so they typically lack flexible attribute models per product type, per-channel completeness rules, rich text, translation workflows and syndication to marketplaces. Businesses with small ranges and one channel can manage in ERP; most others find the effort of customising ERP exceeds the cost of a PIM.
Q: Which system creates a new product?
A: In most organisations ERP or PLM creates the item, because purchasing, costing and logistics need the record before goods can be ordered. The PIM then receives the item and adds commercial content. Some retailers and brands reverse this, creating products in PIM first for early range planning and pushing a minimal record to ERP once the product is approved.
Q: Should prices be managed in the PIM?
A: Transactional prices, discounts and customer-specific pricing should stay in ERP or a dedicated pricing system, because they drive invoicing and margin. A PIM may hold reference prices such as recommended retail price for catalogues or feeds, but it should receive them from the pricing system of record and avoid editing them locally.

## Article: PIM and PLM: handing product data from engineering to commerce
URL: https://pimatlas.com/knowledge/pim-vs-plm/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: PLM (product lifecycle management) governs how a product is designed, specified and changed, from concept to end of life. PIM governs how that product is described and sold across channels. The handover happens when engineering releases a product: PLM sends approved technical data to ERP and PIM, and PIM turns it into channel-ready content.

PLM and PIM cover two consecutive stages of the same product. PLM (product lifecycle management) governs how a product is designed, specified and changed. PIM (product information management) governs how it is described and sold. The handover point is the engineering release: PLM sends approved technical data to ERP and PIM, and PIM turns it into complete, channel-ready content for customers, retailers and dealers.

## What does PLM manage?

CIMdata, the PLM research and advisory firm, defines PLM as "a strategic business approach that applies a consistent set of business solutions that support the collaborative creation, management, dissemination, and use of product definition information" across the extended enterprise, "from concept to end of life". The key phrase is product definition. PLM holds the data that defines what a product is and how it is made.

In manufacturing, that includes CAD models, bills of materials, engineering specifications, change orders and test results. In fashion and consumer goods, retail PLM holds design briefs, materials, colourways, size specifications and supplier samples.

## How do PIM and PLM compare?

| Dimension | PLM | PIM |
|---|---|---|
| Central question | How is this product designed and built? | How is this product described and sold? |
| Main users | Engineering, R&D, quality, regulatory, sourcing | E-commerce, marketing, product data, sales |
| Core objects | Parts, assemblies, BOMs, documents, change orders | Sellable products, variants, families, channel views |
| Typical data | CAD, tolerances, materials, test data, revisions | Copy, marketing attributes, translations, classifications |
| Change control | Formal engineering change process | Enrichment and approval workflow |
| Audience of output | Internal teams, suppliers, manufacturing | Customers, retailers, marketplaces, dealers |
| Lifecycle focus | Concept to end of life | Launch to discontinuation in channels |

## Which data moves from PLM to PIM?

Only a subset of PLM data is relevant for commerce. A useful filter is whether a customer, installer or regulator needs to see it. Data that typically moves includes:

- **Customer-facing specifications:** dimensions, power ratings, materials, performance values.
- **Variant definitions:** which configurations, colours or sizes exist as sellable items.
- **Compliance information:** certifications, declarations of conformity, safety warnings and substance data.
- **Technical documents:** data sheets, installation manuals, safety data sheets, usually stored in DAM and linked from PIM.
- **Lifecycle status:** released, phased out, replaced by, so channels stay current.

Data that should stay in PLM includes CAD geometry, internal tolerances, supplier costings and unreleased revisions. Keeping this boundary tight protects intellectual property and stops draft values from reaching customers before engineering has approved them.

PIM then maps engineering values to commercial structures. A technical value such as "IP65" becomes a classified attribute under a standard such as ETIM, which ETIM International describes as "a format to share and exchange product data based on taxonomic identification". The [manufacturers playbook](https://pimatlas.com/knowledge/pim-for-manufacturers/) covers that mapping in more detail.

## When should the handover happen?

Timing is the most common source of friction. Three models are in use:

1. **Release-triggered.** PLM sends data when a product reaches released status. This is simple and reliable but leaves marketing little time before launch.
2. **Staged.** PLM sends provisional data at an agreed milestone, marked as draft in PIM, then confirms it at release. This supports parallel launch preparation.
3. **Change-driven.** After launch, engineering change orders that affect customer-facing attributes trigger automatic PIM updates and re-approval.

Most manufacturers combine the staged and change-driven models. The [integration architecture article](https://pimatlas.com/knowledge/pim-integration-architecture/) describes the interface options, and the [PIM and ERP article](https://pimatlas.com/knowledge/pim-vs-erp/) explains how ERP fits between the two.

## Why is the PLM to PIM link becoming more important?

Regulation is pushing engineering data towards the customer. The Ecodesign for Sustainable Products Regulation (Regulation (EU) 2024/1781), which entered into force on 18 July 2024, introduces the Digital Product Passport. GS1 in Europe lists priority product groups including iron and steel, aluminium, textiles, furniture, tyres, detergents, paints, chemicals, and ICT and electronics. Passport data such as materials and repairability originates in PLM and must reach customers through digital channels, a topic covered in the [Digital Product Passport article](https://pimatlas.com/knowledge/digital-product-passport/).

## Key takeaways

- PLM manages the engineering definition of a product; PIM manages its commercial description for channels.
- Move only customer-relevant data from PLM to PIM: specifications, variants, compliance data, documents and lifecycle status.
- Choose a handover model deliberately; staged plus change-driven transfers suit most manufacturers.
- Map engineering values to commercial classifications such as ETIM in PIM.
- The ESPR Digital Product Passport makes a reliable PLM to PIM link a compliance requirement for many product groups.

Q: What is the difference between PIM and PLM?
A: PLM manages the engineering definition of a product: CAD models, bills of materials, specifications, change orders and compliance documents throughout its lifecycle. PIM manages the commercial description of a sellable product: marketing content, channel attributes, translations and classifications. PLM serves engineering, quality and regulatory teams; PIM serves e-commerce, marketing and sales teams.
Q: When should product data move from PLM to PIM?
A: Most organisations transfer data when engineering releases a product or variant for sale, often triggered by a lifecycle status change in PLM. Earlier, controlled transfers help when marketing must prepare launch content in parallel with final engineering. After launch, engineering change orders that affect customer-facing specifications should update PIM automatically.
Q: Do retailers and brands without engineering teams need PLM?
A: Fashion, home and food retailers often use PLM to manage design briefs, supplier specifications, materials and samples, even when production is outsourced. These retail PLM systems feed PIM in the same way as engineering PLM in manufacturing, supplying materials, care instructions, composition and sizing that later appear on product pages.

## Article: What is PXM? Product experience management and its relationship to PIM
URL: https://pimatlas.com/knowledge/what-is-pxm/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: PXM (product experience management) is the practice of tailoring, publishing and measuring product content for each channel, audience and context where a product is bought. It builds on PIM: PIM provides the governed product data, and PXM adds channel-specific content, syndication, digital shelf analytics and feedback loops so that content can be improved based on performance.

PXM (product experience management) is the practice of tailoring, publishing and measuring product content for each channel, audience and buying context. It builds directly on PIM (product information management). PIM provides governed, complete product data; PXM adds channel-specific content, syndication to retailers and marketplaces, and analytics that show how the content performs. In short, PXM is PIM plus the capabilities needed to act on what happens after publication.

## Where did the term PXM come from?

PXM is a market label, with no standards body or formal definition behind it. It gained currency among PIM vendors in the late 2010s. In March 2018, Salsify announced what it called the first "Product Experience Management" platform, combining product information management, syndication and e-commerce analytics. In September 2019, Akeneo raised USD 46 million and described itself as a provider of PXM solutions. In October 2022, Akeneo announced Akeneo Product Cloud, which combined PIM with native product asset management and syndication capabilities and an App Store.

Akeneo defines product experience as "every interaction an individual has with your product information". That framing explains the shift: the focus moves from the internal record to the customer's encounter with it.

## How does PXM differ from PIM?

| Capability | Core PIM | PXM suite |
|---|---|---|
| Central product data model | Yes | Yes |
| Enrichment, translation, workflow | Yes | Yes |
| Completeness rules per channel | Yes | Yes |
| Channel-specific content variants (titles, copy, images per retailer) | Partial | Yes |
| Syndication to retailer and marketplace requirements | Often via add-ons or connectors | Yes, usually with maintained retailer templates |
| Supplier and brand content onboarding | Varies | Often included |
| Digital shelf analytics (live listing checks, search position, reviews) | Rare | Common |
| Feedback loop from performance data to content tasks | Rare | Intended core feature |

The boundary is fluid. Many PIM products now include syndication, and some vendors that use the PXM label offer analytics only through partners. Treat PXM as a list of capabilities to verify, independent of the marketing term. A PIM with strong syndication connectors and a separate digital shelf analytics tool can deliver the same outcome as a single PXM suite, at the cost of one more integration and contract.

## What are the building blocks of PXM?

A complete PXM approach rests on four layers:

1. **Governed product data.** A PIM with clear ownership, data models and quality rules. Without it, every downstream layer amplifies errors. See [What is PIM](https://pimatlas.com/knowledge/what-is-pim/).
2. **Contextual content.** Variants of titles, descriptions and images adapted to each channel, language and audience, for example shorter titles for mobile marketplaces or technical depth for B2B portals.
3. **Syndication.** Mapping and delivery of content to each retailer's or marketplace's attribute model and validation rules, explained in the [syndication article](https://pimatlas.com/knowledge/product-data-syndication/).
4. **Measurement and feedback.** Checks that listings are live and correct, plus performance data that turns into enrichment tasks.

## Which organisations benefit most from PXM?

PXM capabilities pay off where the business does not control the final point of sale. Brands that sell through many retailers and marketplaces fit that profile best, because each retailer imposes its own data requirements and presents products differently. Manufacturers selling through distributors face a similar challenge with B2B catalogues.

Retailers that own their storefront usually need less syndication and more on-site personalisation, so their requirements lean towards PIM plus commerce platform features. The rise of AI shopping assistants adds a new context: content must also be structured for machines that summarise products, a topic covered in the [GEO for product content article](https://pimatlas.com/knowledge/geo-product-content/).

## How should buyers evaluate PXM claims?

Ask vendors to demonstrate each layer with your own data. Useful tests include:

- Publish one product to two retailers with different attribute models and show the mapping rules that convert one master record into both formats.
- Show how a retailer's changed requirement reaches users as a task.
- Show a live listing check and how a detected error links back to the product record.
- Clarify which analytics are native, which come from partners, and which require separate licences.

The [selection process article](https://pimatlas.com/knowledge/pim-selection-process/) explains how to script these demonstrations.

## Key takeaways

- PXM is a market term for PIM extended with channel-specific content, syndication and performance measurement.
- Salsify (2018) and Akeneo (2019) were early, prominent users of the PXM label; no formal standard defines it.
- Evaluate PXM as a set of capabilities: governed data, contextual content, syndication and feedback.
- Brands and manufacturers selling through third-party channels gain the most from PXM features.
- Verify in demonstrations which PXM functions are native and which depend on partners or extra licences.

Q: Is PXM the same as PIM?
A: PXM includes PIM and extends it. PIM centralises and governs product data. PXM adds the capabilities that adapt that data for each channel and measure its effect: syndication to retailers and marketplaces, channel-specific content variants, and digital shelf analytics. Many vendors now use PXM to describe suites that combine these functions, so check which modules a PXM offer actually includes.
Q: Who introduced the term product experience management?
A: Several PIM vendors adopted the term in the late 2010s. In March 2018 Salsify announced what it described as the first product experience management platform, combining PIM, syndication and e-commerce analytics. Akeneo described itself as a PXM provider when it raised USD 46 million in September 2019. The term is a vendor and market label, with no formal standard behind it.
Q: What does a PXM platform measure?
A: Typical PXM measurements include content completeness and compliance against each retailer's requirements, whether listings are live and correct on retailer sites, search position for target keywords, ratings and reviews, and the relationship between content changes and conversion or sales. Exact metrics depend on the data each retailer or marketplace makes available.

## Article: Building the business case for PIM: costs, benefits and an ROI model
URL: https://pimatlas.com/knowledge/pim-business-case/
Category: PIM fundamentals | Updated: 2026-10-17

Summary: A credible PIM business case compares five-year total cost of ownership with three measurable benefit streams: content team time saved, revenue from faster product launches, and lower costs from content-related returns and errors. Use your own baseline data, show every assumption, and test how sensitive the result is to the revenue assumption, which usually decides the outcome.

A credible PIM business case sets five-year total cost of ownership against three measurable benefit streams: time saved in content work, revenue from launching products sooner, and lower costs from content-related returns and errors. It rests on your own baseline data, shows every assumption in plain view and tests which assumption the result depends on. The worked example below uses illustrative figures for a mid-sized manufacturer; replace each one with your own.

## Which costs should the business case include?

Most PIM business cases understate cost because they stop at the software fee. A complete view over five years covers one-off and recurring items.

**One-off costs**
- Implementation partner fees for configuration, data modelling and integrations.
- Data cleansing, enrichment backlog and migration from spreadsheets or legacy systems.
- Internal project team time, including product data owners and IT.
- Training and change management.

**Recurring costs**
- Subscription or licence fees, often tiered by SKUs, users, channels or locales.
- Connector and syndication fees for marketplaces and retailers.
- Support, hosting (for on-premise or private cloud), and a product data administrator role.

The [pricing models article](https://pimatlas.com/knowledge/pim-pricing-models/) explains how vendors charge, and the [total cost of ownership article](https://pimatlas.com/knowledge/pim-total-cost-of-ownership/) covers the five-year view in detail.

## Which benefits can be measured?

Limit the case to benefits you can baseline today and measure after go-live.

1. **Content team productivity.** Hours spent collecting, correcting, translating and reformatting product data for each channel.
2. **Faster time to market.** The number of weeks between a product being available to ship and being fully listed on all channels, multiplied by the margin lost in that gap.
3. **Lower cost of errors.** Returns and complaints caused by wrong or missing product information, plus listing rejections from retailers and marketplaces.

Compliance is a fourth driver that rarely fits an ROI formula. The EU General Product Safety Regulation (Regulation (EU) 2023/988) has applied since 13 December 2024 and requires online offers to show manufacturer details, product identification including a picture, and warnings or safety information. Present that as risk reduction alongside the financial model.

## A worked example: five-year ROI model

All figures below are **illustrative assumptions** for a manufacturer with 20,000 SKUs, six content staff and 2,000 new products a year. They are not industry benchmarks.

| Assumption (illustrative) | Value |
|---|---|
| Content staff, fully loaded cost each | 6 × €65,000 |
| Share of their time saved by PIM | 30% |
| New SKUs launched per year | 2,000 |
| Weeks faster to full listing | 2 |
| Gross margin per new SKU per week | €35 |
| Annual cost of content-related returns | €200,000 |
| Reduction in those returns | 25% |
| One-off costs (implementation €250,000, data work €100,000) | €350,000 |
| Recurring costs (subscription €90,000, support and admin €30,000) | €120,000 per year |

Full annual benefit = labour €117,000 + faster launches €140,000 + returns €50,000 = **€307,000**. Benefits ramp up at 25% in year one and 75% in year two, reaching 100% from year three.

| Year | Benefits | Costs | Net | Cumulative |
|---|---|---|---|---|
| 1 | €76,750 | €470,000 | −€393,250 | −€393,250 |
| 2 | €230,250 | €120,000 | €110,250 | −€283,000 |
| 3 | €307,000 | €120,000 | €187,000 | −€96,000 |
| 4 | €307,000 | €120,000 | €187,000 | €91,000 |
| 5 | €307,000 | €120,000 | €187,000 | €278,000 |

Over five years, benefits of €1,228,000 against costs of €950,000 give a net return of €278,000, an undiscounted ROI of 29%, with payback during year four. Apply your organisation's discount rate if finance requires net present value.

## How sensitive is the result?

The model depends heavily on the time-to-market assumption. Halve the margin gained from faster launches and the five-year result falls to roughly break-even. Remove it entirely and the case shows a net cost of €282,000. That tells the board exactly which assumption to validate, for example by measuring current listing lead times on a sample of recent launches.

Run the same test on every major input. Assumptions that swing the result should be measured before approval, and the [selection process](https://pimatlas.com/knowledge/pim-selection-process/) should confirm that shortlisted vendors can support them.

## Key takeaways

- Count five years of one-off and recurring costs, including data work, integrations and internal time.
- Restrict benefits to streams you can baseline now: content productivity, time to market and content-related errors.
- Label every input as an assumption and replace illustrative values with your own data.
- In the example, a €307,000 annual benefit pays back during year four with a 29% five-year ROI.
- Run sensitivity tests; the time-to-market assumption often decides whether the case is positive.

Q: What ROI can a PIM deliver?
A: Returns depend on range size, channel count, current processes and implementation quality, so no general figure is reliable. Vendor-commissioned studies report high returns for composite organisations, which may differ from your situation. Build the case from your own baseline: hours spent on product content, time to market for new items, return rates linked to content, and channel listing errors.
Q: What costs belong in a PIM business case?
A: Include software subscription or licence fees, implementation partner fees, data cleansing and migration, integration with ERP, DAM and channels, internal project staff time, training, and ongoing support, hosting and administration. Budget for connector or syndication fees where channels require them. Evaluate all costs over at least five years to compare SaaS and on-premise options fairly.
Q: How long does a PIM take to pay back?
A: Payback depends on the size of benefits relative to one-off costs and on how quickly teams adopt the system. In the illustrative model in this article, cumulative benefits overtake cumulative costs during year four. Phasing the rollout so the highest-value channel or product range goes live first is the most practical way to shorten payback.

## Article: How to select a PIM system: an eight-step process
URL: https://pimatlas.com/knowledge/pim-selection-process/
Category: Selection and buying | Updated: 2026-10-17

Summary: A sound PIM selection runs in eight steps: define scope and outcomes, map current product data, write weighted requirements, build a longlist, issue an RFP, run scripted demos with your own data, test the finalist in a proof of concept, then negotiate on five-year total cost of ownership. Most mid-market selections take 12 to 20 weeks.

A PIM (product information management) system is selected most reliably through eight sequential steps that move from business outcomes to contract. Each step narrows the field with evidence: documented requirements, written RFP answers, scripted demos with your own data and a proof of concept.
## What are the eight steps of a PIM selection?

The table below sets out each step with its output and a typical duration for a mid-market organisation with 10,000 to 200,000 SKUs.

| Step | Activity | Output | Typical duration |
|---|---|---|---|
| 1 | Define scope and outcomes | Signed-off goals, KPIs, budget range | 1 to 2 weeks |
| 2 | Map current product data and systems | Data inventory, integration map, pain points | 2 to 3 weeks |
| 3 | Write weighted requirements | Requirements catalogue with MoSCoW priority and weights | 2 weeks |
| 4 | Build a longlist | 8 to 12 vendors screened on fit criteria | 1 week |
| 5 | Issue an RFP | 4 to 6 scored written responses | 3 to 4 weeks |
| 6 | Run scripted demos | 3 vendors scored on your scenarios | 2 to 3 weeks |
| 7 | Proof of concept | 1 or 2 finalists tested on real data | 2 to 4 weeks |
| 8 | Negotiate on total cost of ownership | Contract, SLA, exit terms, implementation plan | 2 to 4 weeks |

Steps 2 and 3 can overlap, and step 4 can start once the scope is stable. Teams that skip step 2 usually discover integration and data quality problems during implementation, when changes cost the most.

## How do you define scope and requirements?

Start with outcomes before features. Agree three to five measurable goals, such as time to market for new products, completeness scores per channel, or the number of marketplaces served. Tie each goal to the [PIM business case](https://pimatlas.com/knowledge/pim-business-case/) so the budget conversation rests on expected benefits.

Then document the current state: where product data originates (ERP, PLM, supplier files, spreadsheets), how many attributes and locales exist, which channels consume the data, and who edits what. This inventory feeds the requirements catalogue. Write requirements as testable statements and weight them. A buyer who writes "supports variants" learns little; "manages three-level variants (model, colour, size) with attribute inheritance" separates vendors. The [PIM RFP requirements guide](https://pimatlas.com/knowledge/pim-rfp-requirements/) covers the categories that matter most.

## How do you build a longlist and shortlist?

Screen the market on hard constraints first. These knock-out criteria remove vendors quickly and fairly:

- **Deployment model**: SaaS only, private cloud or on-premise, and data residency in the EU if required.
- **Scale**: SKU count, attribute count, locales and expected growth over five years.
- **Industry fit**: references in your sector, support for ETIM, GS1 GDSN or other standards you must publish.
- **Integration**: native or proven connectors for your ERP, commerce platform and DAM.
- **Budget band**: published or indicative pricing within your approved range.
- **Partner ecosystem**: certified implementation partners in your country and language.

Use the [vendor directory](https://pimatlas.com/vendors/) and [compare tool](https://pimatlas.com/compare/) to filter on these criteria, then send the RFP to the four to six vendors that pass.

## How should demos and a proof of concept be run?

Scripted demos are the step where most selections are won or lost. Give each vendor the same scenarios and a sample of your own product data two weeks in advance, then score every vendor against the same rubric. Unscripted demos favour the vendor with the best presenter. The [guide to scripted PIM vendor demos](https://pimatlas.com/knowledge/pim-vendor-demos/) sets out a full day structure and scoring model.

A proof of concept follows for the finalist or the final two. Limit it to two to four weeks and three to five critical scenarios, such as importing a real supplier file, modelling your most complex product family and publishing to one channel. Agree in writing who pays for it. Some vendors fund a short proof of concept; others charge a fixed fee that is credited against the first year's licence.

## How do you make the final decision?

Score the finalists on a weighted model that combines functional fit, technical fit, vendor viability, implementation approach and five-year cost. A common split is 40% functional, 20% technical, 15% vendor and partner, 25% commercial. Quality frameworks such as ISO/IEC 25010:2023, which defines nine product quality characteristics, help structure the non-functional criteria such as performance, security and maintainability.

Compare cost over five years, including implementation, integrations and internal staff, using a [total cost of ownership model](https://pimatlas.com/knowledge/pim-total-cost-of-ownership/). Public-sector buyers in the EU must also state the relative weighting of award criteria in the procurement documents under Article 67(5) of Directive 2014/24/EU, so fix the weights before the RFP is issued. Private buyers gain from the same discipline.

## Key takeaways

- Run eight steps in order: scope, data mapping, requirements, longlist, RFP, scripted demos, proof of concept, negotiation.
- Write testable, weighted requirements and fix the scoring weights before any vendor contact.
- Use knock-out criteria (deployment, scale, standards, integrations, budget) to cut the longlist quickly.
- Score scripted demos with your own data and prove the riskiest scenarios in a short proof of concept.
- Decide on five-year total cost of ownership and contract terms, including exit and data portability.

Q: How long does a PIM selection take?
A: A structured PIM selection for a mid-market organisation typically takes 12 to 20 weeks from kick-off to signed contract. Smaller teams with a simple catalogue can finish in 8 weeks, while enterprises with formal procurement, multiple business units and a proof of concept often need six months. The longest delays usually come from unclear scope and slow internal sign-off, so fix both early.
Q: How many PIM vendors should be on a shortlist?
A: Keep the longlist to 8 to 12 vendors and the RFP shortlist to 4 to 6. Invite 3 vendors to scripted demos and take 1 or 2 into a proof of concept. Larger shortlists dilute evaluator attention and slow scoring, while fewer than three demos leaves no meaningful comparison and weakens your position in commercial negotiation.
Q: Who should be on a PIM selection team?
A: A typical selection team includes an executive sponsor, a product data or e-commerce lead as owner, an IT or enterprise architect, two or three daily users from category management or marketing, and a procurement representative. Add legal and information security for contract and hosting review. Every scorer should attend every demo so scores are comparable across vendors.

## Article: Writing a PIM RFP: the requirements that separate vendors
URL: https://pimatlas.com/knowledge/pim-rfp-requirements/
Category: Selection and buying | Updated: 2026-10-17

Summary: A useful PIM RFP asks fewer, sharper questions. Focus on eight requirement areas where vendors genuinely differ: data model and variants, data quality, workflow, integration and APIs, syndication and standards, AI, scale and performance, and commercial terms. Write each requirement as a testable scenario, weight it in advance and ask for evidence.

A PIM RFP (request for proposal) separates vendors only when its requirements are specific, testable and weighted before vendors see them. Most PIM platforms can store attributes, import spreadsheets and expose an API, so generic questions produce identical "yes" answers. The requirements that separate vendors describe your hardest products, your channels and your governance rules in concrete terms.

## Which requirement areas separate PIM vendors?

The table lists eight areas where answers differ materially between vendors, with example requirements that force a specific answer.

| Area | Example requirement that separates vendors | Suggested weight |
|---|---|---|
| Data model and variants | Model three variant levels with inheritance and override per channel | 15% |
| Data quality | Completeness rules per channel and locale, with blocking validation on publish | 10% |
| Workflow and governance | Configurable approval steps by attribute group and role, with audit trail | 10% |
| Integration and APIs | Bulk REST or GraphQL API, event webhooks, documented rate limits | 15% |
| Syndication and standards | Export to GS1 GDSN, ETIM or BMEcat, plus named marketplaces | 15% |
| AI capabilities | Attribute extraction and translation with human review and audit log | 10% |
| Scale and performance | Import 500,000 SKUs with 300 attributes within a stated time | 10% |
| Commercial and contract | Five-year pricing, renewal caps, exit and data export terms | 15% |

Adjust the weights to your situation. A distributor onboarding thousands of supplier files will weight imports and data quality higher; a brand selling on 20 marketplaces will weight syndication higher. Fix the weights before the RFP is issued, as public buyers must under Article 67(5) of Directive 2014/24/EU.

## How should PIM requirements be written?

Write every requirement as a scenario with a measurable result. Each one should say who does what, with which data, and what the system must produce. Compare these two versions:

- Weak: "The system supports multiple languages."
- Strong: "A translator edits French and German values for 40 marketing attributes, while technical attributes inherit from English and stay locked."

Ask vendors to classify each answer: standard (configuration only), extension (marketplace module or paid add-on), custom (development required) or roadmap (with a date). This classification matters more than a yes or no, because custom and roadmap answers carry cost and delivery risk. Request evidence for critical items: documentation links, screenshots or a short video.

Base the data model section on your real product families. The [product data modelling guide](https://pimatlas.com/knowledge/product-data-modelling/) explains families, attributes and variants in detail, and the [integration architecture guide](https://pimatlas.com/knowledge/pim-integration-architecture/) covers the API and middleware questions worth asking.

## What non-functional requirements belong in a PIM RFP?

Non-functional requirements often decide long-term success. ISO/IEC 25010:2023 provides a recognised structure for product quality, and the following items deserve explicit questions:

1. **Hosting and residency**: data centre locations, EU hosting options and sub-processors.
2. **Security**: certifications held (for example ISO/IEC 27001 or SOC 2), single sign-on, role-based access.
3. **Availability**: SLA percentage, measurement method and service credits.
4. **Environments**: number of sandbox and staging environments included.
5. **Release policy**: frequency of updates and notice periods for breaking API changes.
6. **Exit**: export formats, retrieval period and switching support. The EU Data Act, Regulation (EU) 2023/2854, sets switching and data-retrieval rights for cloud services that apply from 12 September 2025.

## What should a PIM RFP leave out?

Long RFPs reduce answer quality. Leave out requirements every vendor meets, such as "users can log in" or "data can be exported to CSV". Avoid questions that invite marketing text, such as "describe your vision". Do not prescribe a technical design; describe the outcome and let vendors explain how they reach it. Remove duplicate questions across sections, which often appear when several departments contribute.

Keep the document to 80 to 150 requirements and a pricing template. Give vendors three to four weeks to respond and a single question deadline, with answers shared with all bidders.

## How do you score RFP responses?

Use a 0 to 3 or 0 to 5 scale per requirement, with written definitions for each score. Multiply by weight and sum per area. Score independently first, then hold a calibration session to resolve differences larger than one point. Shortlist the top three for [scripted demos](https://pimatlas.com/knowledge/pim-vendor-demos/), where you verify the written claims against your own data. The overall flow is described in the [eight-step PIM selection process](https://pimatlas.com/knowledge/pim-selection-process/).

## Key takeaways

- Focus the RFP on eight areas where vendors differ: data model, quality, workflow, APIs, standards, AI, scale and commercial terms.
- Write each requirement as a testable scenario based on your real products and channels.
- Ask vendors to classify answers as standard, extension, custom or roadmap, and request evidence.
- Include non-functional and exit requirements, referencing the EU Data Act switching rules.
- Fix weights and scoring definitions before issue, then verify written answers in scripted demos.

Q: How many requirements should a PIM RFP contain?
A: Most effective PIM RFPs contain 80 to 150 requirements. Below 50, the document rarely captures enough detail to separate vendors. Above 250, vendors answer with copied marketing text and evaluators lose focus. Concentrate the detail on the areas specific to your business, such as variant structures, standards you must publish and integrations, and keep generic items short.
Q: Should a PIM RFP include pricing questions?
A: Yes. Ask for pricing in a fixed template covering licence or subscription by year for five years, the pricing metric (SKUs, users, channels, environments), included and paid modules, implementation estimates, support tiers and renewal caps. A common template makes offers comparable and exposes costs that would otherwise appear only after contract signature, such as extra environments or API call limits.
Q: What is the difference between an RFI and an RFP for PIM?
A: An RFI (request for information) is a short questionnaire used to screen a longlist of 8 to 12 vendors on basic fit, such as deployment model, scale and industry references. An RFP (request for proposal) goes to a shortlist of 4 to 6 vendors and asks for detailed answers, pricing and an implementation approach that you score against weighted requirements.

## Article: How much does a PIM cost? Pricing models explained
URL: https://pimatlas.com/knowledge/pim-pricing-models/
Category: Selection and buying | Updated: 2026-10-17

Summary: PIM software costs range from free self-hosted open source editions to six-figure annual enterprise subscriptions. Published 2026 entry prices start around EUR 149 per month for SMB SaaS and roughly USD 10,000 to 30,000 per year for commercial editions, while enterprise platforms are quote-based. Pricing usually scales with SKUs, users, channels, environments or modules.

A PIM (product information management) system can cost anything from nothing for a self-hosted open source edition to several hundred thousand euros a year for an enterprise subscription. The price depends less on the vendor's brand than on the pricing metric: SKUs, users, channels, environments or modules. Licence fees are also only part of the spend, since implementation and integration often equal or exceed the first year's subscription.

## What pricing models do PIM vendors use?

Five pricing metrics appear across the market, often in combination.

| Model | How it works | Watch for |
|---|---|---|
| SKU tiers | Price rises with the number of products or variants | Whether variants count as separate SKUs |
| Users or seats | Price per named or concurrent editor | Read-only and supplier users counted as seats |
| Edition and modules | Base edition plus paid add-ons (DAM, syndication, AI) | Core features sold as add-ons |
| Channels and connectors | Fee per marketplace, feed or connector | Cost of each new channel over five years |
| Platform and environments | Flat platform fee plus sandbox or production instances | Extra environments, API call limits, storage |

Most SMB SaaS tools price by SKU band and include unlimited users. Enterprise vendors usually combine a platform fee with tiers for SKUs, users and modules, then discount by contract length and volume. Open source vendors typically charge for hosting, support and premium modules instead of the core licence, as described in the [open source PIM guide](https://pimatlas.com/knowledge/open-source-pim/).

## What do published PIM prices look like in 2026?

A minority of vendors publish prices. The examples below were taken from vendor pricing pages in October 2026 and show starting points only.

- **[Plytix](https://pimatlas.com/vendors/plytix/)**: Standard from EUR 0 per month for 100 SKUs, Pro from EUR 149 per month and Enterprise from EUR 699 per month, rising with SKU count; add-ons such as feed management are priced separately.
- **[Ergonode](https://pimatlas.com/vendors/ergonode/)**: a free plan limited to 1,000 products and 2 users, Advance from EUR 5,990 per year and Scale from EUR 19,990 per year, with Enterprise on quote.
- **[Pimcore](https://pimatlas.com/vendors/pimcore/)**: Professional Edition at USD 9,900 per year and Enterprise Edition at USD 29,900 per year for on-premise use, with PaaS on quote and no stated limits on products, assets or seats.
- **[AtroPIM](https://pimatlas.com/vendors/atropim/)**: a free open source edition for self-hosting; the vendor's SaaS edition starts from EUR 400 per month, with premium modules and integrations charged separately.

Most enterprise platforms do not publish prices and quote after a discovery call. Treat any figure found on third-party sites with caution, since list prices, discounts and bundles change frequently. Published prices also exclude VAT, implementation partner fees and the internal team that maintains the data model after go-live.

## How much should you budget for a PIM?

The ranges below are illustrative, based on typical deal structures, and are not market statistics. Use them only to sanity-check vendor offers.

| Segment | Catalogue size | Annual subscription (illustrative) | One-off implementation (illustrative) |
|---|---|---|---|
| Small business | Under 10,000 SKUs | EUR 0 to 15,000 | EUR 5,000 to 30,000 |
| Mid-market | 10,000 to 250,000 SKUs | EUR 20,000 to 80,000 | EUR 50,000 to 250,000 |
| Enterprise | Above 250,000 SKUs, many locales | EUR 80,000 to 300,000+ | EUR 250,000 to 1,000,000+ |

Implementation cost depends mainly on data model complexity, the number of integrations and the state of source data. The [total cost of ownership guide](https://pimatlas.com/knowledge/pim-total-cost-of-ownership/) provides a five-year model that includes internal staff and run costs.

## Which cost drivers catch buyers out?

Several items regularly appear after signature:

1. Variant counting rules that multiply the SKU tier.
2. Paid connectors for each marketplace or commerce platform.
3. Extra sandbox environments needed for testing and training.
4. API rate limits that require an upgrade once integrations go live.
5. AI credits consumed by enrichment and translation features.
6. Annual uplifts at renewal without a contractual cap.
7. Premium support tiers required for a meaningful SLA.

Ask each shortlisted vendor to complete the same five-year pricing template in the RFP, stating the metric, included volumes, overage rates and renewal caps.

## Key takeaways

- PIM prices depend on the metric: SKUs, users, modules, channels or environments.
- Published 2026 starting prices range from free plans to USD 29,900 per year; enterprise pricing is quote-based.
- Implementation often equals or exceeds the first-year subscription.
- Variant counting, connectors, environments, API limits and renewal uplifts are the most common hidden costs.
- Compare vendors on a common five-year template that includes overage rates and renewal caps.

Q: How much does a PIM cost per year?
A: Published 2026 prices show SMB SaaS plans from about EUR 1,800 per year (Plytix Pro from EUR 149 per month), commercial editions such as Pimcore Professional at USD 9,900 per year, and mid-tier SaaS plans around EUR 20,000 per year. Enterprise platforms are quote-based and, as an illustrative range, often cost EUR 50,000 to 250,000 or more per year before implementation.
Q: Why do most PIM vendors not publish prices?
A: Enterprise PIM vendors price on several variables at once, including SKU volume, number of users, locales, channels, environments, modules and support level, and they discount by deal size and contract length. A single list price would mislead most buyers. Ask every shortlisted vendor to complete the same five-year pricing template so offers can be compared directly.
Q: Is PIM software priced per SKU or per user?
A: Both models exist and many vendors combine them. SMB SaaS tools often price by SKU band with unlimited users, while enterprise platforms tend to combine a platform fee with tiers for SKUs, users, channels or modules. Some open source vendors charge only for hosting, support and premium modules. Check which metric grows fastest in your business before signing.

## Article: SaaS, private cloud or on-premise PIM: choosing a deployment model
URL: https://pimatlas.com/knowledge/saas-vs-on-premise-pim/
Category: Selection and buying | Updated: 2026-10-17

Summary: Multi-tenant SaaS is the default PIM deployment model in 2026 because it removes infrastructure work and delivers continuous updates. Private cloud suits organisations that need dedicated infrastructure, custom code or specific residency. On-premise or self-hosted PIM fits strict security policies, heavy customisation or open source strategies, at the cost of owning upgrades and operations.

Multi-tenant SaaS (software as a service) is the default choice for most new PIM (product information management) projects in 2026. It shifts hosting, patching and upgrades to the vendor and gives predictable subscription costs. Private cloud and on-premise remain valid when security policy, customisation depth or an open source strategy requires direct control, and buyers accept the extra operational work.

## What are the three PIM deployment models?

The NIST definition of cloud computing (SP 800-145) separates service models such as SaaS from deployment models such as private and public cloud. For PIM buyers, three practical options follow from that distinction.

- **Multi-tenant SaaS**: the vendor runs one application for many customers. Updates arrive continuously and customers extend the system through APIs, configuration and marketplace apps.
- **Private cloud (single-tenant or PaaS)**: the vendor or a partner runs a dedicated instance for one customer. Custom code is possible and upgrade timing can be negotiated.
- **On-premise or self-hosted**: the customer installs and operates the software on its own servers or its own cloud account, and owns upgrades, backups and security.

## How do the models compare?

| Criterion | Multi-tenant SaaS | Private cloud | On-premise or self-hosted |
|---|---|---|---|
| Infrastructure work | None | Low (vendor or partner runs it) | High (own team) |
| Upgrades | Continuous, automatic | Scheduled, negotiated | Own project, own timing |
| Customisation | Configuration, APIs, apps | Custom code allowed | Full code access |
| Data residency control | Vendor's regions | Chosen region | Full control |
| Cost profile | Subscription, operating expense | Higher subscription | Licence or free core plus staff and hosting |
| Time to first environment | Days | Weeks | Weeks |
| Exit effort | Export and migrate | Export and migrate | Data already in-house |

The cost profile matters for budgeting. SaaS bundles hosting and operations into the subscription. Self-hosting removes the subscription for open source editions but adds servers, monitoring, security patching and developer time. The [PIM pricing guide](https://pimatlas.com/knowledge/pim-pricing-models/) and [total cost of ownership model](https://pimatlas.com/knowledge/pim-total-cost-of-ownership/) show how to compare these on a five-year basis.

## Which vendors still offer on-premise PIM?

Availability changes, so verify current editions before shortlisting. Examples verified in October 2026:

- [Akeneo](https://pimatlas.com/vendors/akeneo/) lists 30 September 2026 as the end of support for version 7.0 of its PaaS and on-premise editions, and focuses commercial development on its SaaS platform. Its Community Edition remains available on GitHub under the Open Software License 3.0.
- [Pimcore](https://pimatlas.com/vendors/pimcore/) publishes on-premise Professional and Enterprise Edition prices and offers a separate PaaS option on quote.
- [AtroPIM](https://pimatlas.com/vendors/atropim/) offers a free self-hosted open source edition and a hosted SaaS edition.

Several enterprise vendors, particularly in product MDM, continue to support customer-managed installations for regulated industries. Ask each vendor for its support policy and end-of-life dates for every edition it proposes.

## Which model fits which organisation?

Use these criteria to decide:

1. **Choose multi-tenant SaaS** when the team has little infrastructure capacity, requirements fit configuration and APIs, and fast time to value matters most.
2. **Choose private cloud** when you need custom code, a fixed hosting region, a dedicated performance profile or controlled upgrade windows tied to release cycles.
3. **Choose on-premise or self-hosted** when security policy forbids external hosting of product data, when deep customisation is central, or when an [open source PIM](https://pimatlas.com/knowledge/open-source-pim/) is part of the strategy and in-house developers are available.

Integration patterns also differ. SaaS favours API-led and event-based integration through middleware, while on-premise installations often still rely on direct database access or file transfers, which complicate a later move to the cloud.

## What should the contract cover for each model?

For SaaS and private cloud, the contract should state hosting regions, sub-processors, SLA percentage, backup frequency, notice periods for breaking changes and exit support. The EU Data Act, Regulation (EU) 2023/2854, gives customers of cloud services switching rights from 12 September 2025: a maximum two-month notice period, a 30-day transition and at least 30 days to retrieve data. Switching charges are limited to the provider's direct costs until 11 January 2027 and are abolished from 12 January 2027.

For on-premise, check licence terms for production use, the support period for each major version and the availability of security patches after end of support.

## Key takeaways

- Multi-tenant SaaS is the default for new PIM projects because it removes infrastructure and upgrade work.
- Private cloud suits custom code, fixed regions or controlled upgrade timing at a higher price.
- On-premise and self-hosted fit strict security policies and open source strategies with in-house skills.
- Verify each vendor's end-of-support dates; Akeneo version 7.0 PaaS and on-premise support ends 30 September 2026.
- Write hosting regions, SLA, exit support and EU Data Act switching rights into the contract.

Q: Is SaaS PIM more secure than on-premise PIM?
A: Neither model is secure by default. SaaS vendors typically hold certifications such as ISO/IEC 27001 or SOC 2 and patch continuously, which many internal IT teams cannot match. On-premise gives full control over network, access and encryption, but security depends entirely on your own team. Compare certifications, penetration test results and patching responsibilities instead of assuming one model is safer.
Q: Can I move from on-premise PIM to SaaS later?
A: Yes, but it is a migration project rather than a configuration switch. Custom code written for an on-premise installation often has to be rebuilt as API-based extensions, and data must be exported, cleansed and reimported. Plan it like a new implementation with its own budget, and check the vendor's documented migration path and tooling before choosing an on-premise edition today.
Q: What is the difference between private cloud and SaaS PIM?
A: In multi-tenant SaaS, many customers share the same application instance and receive identical updates on the vendor's schedule. In private cloud, the vendor or a partner runs a dedicated instance for one customer, which allows more custom code, controlled upgrade timing and specific hosting locations. Private cloud usually costs more and requires more coordination on upgrades.

## Article: Open source PIM: options, licences and trade-offs
URL: https://pimatlas.com/knowledge/open-source-pim/
Category: Selection and buying | Updated: 2026-10-17

Summary: Open source PIM in 2026 means AtroPIM (GPLv3), UnoPim (MIT) and Akeneo Community Edition (OSL 3.0), plus Pimcore, whose Community Edition moved to the Pimcore Open Core License with a EUR 5 million revenue threshold. The licence is free, but hosting, development, upgrades and support usually make up most of the total cost.

Open source PIM (product information management) software gives buyers the source code, the right to self-host and freedom from per-SKU subscription fees. In 2026 the main options are AtroPIM, UnoPim and Akeneo Community Edition, with Pimcore now under an open core licence that charges larger organisations. The trade-off is clear: the licence is free, while operating, upgrading and extending the system is the buyer's responsibility.

## Which open source PIM options exist in 2026?

Licences changed in 2025, so check the current terms for each product. The table reflects documentation and repositories reviewed in October 2026.

| Product | Licence | Technology | Commercial offer |
|---|---|---|---|
| [Akeneo](https://pimatlas.com/vendors/akeneo/) PIM Community Edition | Open Software License 3.0 | PHP, Symfony | SaaS platform (separate product) |
| [Pimcore](https://pimatlas.com/vendors/pimcore/) Community Edition | Pimcore Open Core License (POCL) | PHP, Symfony | Professional, Enterprise, PaaS |
| [AtroPIM](https://pimatlas.com/vendors/atropim/) | GNU GPLv3 | PHP, built on AtroCore | Premium modules, support, SaaS |
| [UnoPim](https://pimatlas.com/vendors/unopim/) | MIT | PHP, Laravel | Hosting, paid connectors, services (Webkul) |
| [Ergonode](https://pimatlas.com/vendors/ergonode/) | OSL 3.0 (archived repository) | PHP, Symfony | Commercial SaaS plans, including a free tier |

Two cases need explanation. Pimcore moved its Community Edition from GPLv3 to POCL with Platform Version 2025.1; it remains free for organisations below EUR 5 million annual revenue, and larger organisations need a commercial licence for production use. Ergonode archived its open source backend repository on 13 February 2023 and now sells hosted plans, so treat it as a commercial SaaS product with a free plan.

## How do open source PIM licences differ?

The licence decides what you may do with modified code.

- **MIT (UnoPim)**: permissive. You may modify, combine and redistribute the code with few obligations beyond keeping the copyright notice.
- **GPLv3 (AtroPIM)**: copyleft. Modified versions you distribute must be released under GPLv3.
- **OSL 3.0 (Akeneo Community Edition)**: copyleft with an "external deployment" clause. Making modified software available to others over a network counts as distribution, which matters if you expose the PIM to suppliers or partners.
- **POCL (Pimcore)**: free below a revenue threshold, with restrictions on offering the core as SaaS without an OEM licence. Treat it as a commercial licence with a free tier.

Ask your legal team to review the licence before writing custom extensions, particularly if partners or suppliers will use the system.

## What does open source PIM really cost?

The licence is only one line of the budget. The model below is illustrative and assumes a mid-market catalogue of about 50,000 SKUs, self-hosted.

| Cost item | Illustrative annual range |
|---|---|
| Cloud hosting, backups, monitoring | EUR 3,000 to 15,000 |
| Security patching and version upgrades | EUR 10,000 to 40,000 |
| Feature development and connectors | EUR 20,000 to 80,000 |
| Professional support subscription (optional) | EUR 3,000 to 10,000 |

Some vendors publish support and module prices. AtroCore lists optional professional support from EUR 240 per month and on-premise premium modules as one-off licences with a 20% annual update fee. Compare these figures with SaaS subscriptions over five years using the [total cost of ownership model](https://pimatlas.com/knowledge/pim-total-cost-of-ownership/).

## When is open source PIM the right choice?

Open source suits organisations that meet most of these conditions:

1. In-house PHP developers or a long-term partner with experience of the chosen framework.
2. A data model or workflow that needs code-level customisation.
3. A policy preference for self-hosting or full control over data location.
4. A catalogue where per-SKU SaaS pricing would scale poorly.
5. Acceptance that upgrades are a recurring project with their own budget.

It suits organisations with no developers, a standard data model and a need for vendor-managed syndication less well. Those buyers usually reach value faster with a multi-tenant SaaS PIM, where hosting and upgrades are included in the subscription.

## What risks should buyers check?

Review four risks before committing. First, project activity: check recent releases, commit frequency and open security issues. Second, support policy: Akeneo states that only the latest Community Edition release receives fixes and that breaking changes may appear in any release. Third, licence changes: Pimcore's 2025 switch shows that terms can change between versions. Fourth, the partner market: confirm that certified implementers exist in your country and language.

## Key takeaways

- The main open source PIM options in 2026 are AtroPIM (GPLv3), UnoPim (MIT) and Akeneo Community Edition (OSL 3.0).
- Pimcore's Community Edition moved to POCL in 2025 and is free only below EUR 5 million annual revenue.
- Ergonode archived its open source repository in 2023 and now sells hosted plans with a free tier.
- Hosting, upgrades, development and support usually exceed what a SaaS subscription would cost for teams without developers.
- Check project activity, support policy, licence terms and partner availability before choosing.

Q: Is Akeneo still open source?
A: Akeneo PIM Community Edition remains available on GitHub under the Open Software License 3.0, and Akeneo tagged a v2026.3 release in March 2026 under calendar versioning. Only the latest release receives fixes, and backwards compatibility is not guaranteed. Akeneo's commercial focus is its SaaS platform, and version 7.0 PaaS and on-premise support ends on 30 September 2026.
Q: Is Pimcore still free to use?
A: Since Pimcore Platform Version 2025.1, the Community Edition is licensed under the Pimcore Open Core License (POCL) instead of GPLv3. It is free for organisations with annual revenue below EUR 5 million. Organisations above that threshold need a commercial licence for production use. Version 2024.4 was the last release under GPLv3.
Q: Is open source PIM cheaper than SaaS PIM?
A: Not automatically. An open source licence removes the subscription fee, but the organisation pays for hosting, security patching, upgrades, development and support, either internally or through a partner. Open source is usually cheaper when developers are already in-house and the catalogue needs heavy customisation, and more expensive when the team would have to hire specialists only for the PIM.

## Article: Running scripted PIM vendor demos that reveal real fit
URL: https://pimatlas.com/knowledge/pim-vendor-demos/
Category: Selection and buying | Updated: 2026-10-17

Summary: A scripted PIM demo gives every shortlisted vendor the same scenarios and a sample of your own product data, then scores each scenario on a shared rubric. Send scripts two weeks ahead, run one structured day per vendor, score individually on a 0 to 3 scale and calibrate together. This exposes real fit far better than a standard sales presentation.

A scripted PIM (product information management) demo is a structured session in which every shortlisted vendor performs the same scenarios on a sample of your own product data, scored against one rubric. It replaces the vendor's standard sales presentation with evidence of how the software handles your products, processes and integrations. Scripted demos are the most reliable way to compare three finalists before a proof of concept.

## Why use scripted demos instead of standard presentations?

Standard demos show prepared data and the vendor's strongest features in the vendor's preferred order. Every product looks capable in that setting. A scripted demo reverses control: the buyer defines the order, the data and the success criteria, so gaps become visible and answers become comparable across vendors.

Scripts also verify written RFP claims. An answer of "standard" in the [PIM RFP](https://pimatlas.com/knowledge/pim-rfp-requirements/) should be demonstrable through configuration on the day, without custom code.

## How do you write demo scenarios?

Write six to eight scenarios that reflect your hardest daily work. Each scenario needs a named persona, input data, the steps expected and a measurable result. Typical scenarios include:

1. **Supplier onboarding**: import a real supplier spreadsheet with errors, map columns, and show which records fail validation and why.
2. **Complex product modelling**: create your most complex family with three variant levels and inherited attributes. The [data modelling guide](https://pimatlas.com/knowledge/product-data-modelling/) explains these structures.
3. **Enrichment and AI**: generate descriptions or translations for 20 products, with human approval and an audit trail.
4. **Data quality**: show completeness per channel and locale and block publication of incomplete products. See [product data quality](https://pimatlas.com/knowledge/product-data-quality/) for useful metrics.
5. **Workflow**: route a new product through approval by category, legal and marketing roles.
6. **Syndication**: publish to one marketplace or a GS1 GDSN data pool and show the error-handling loop.
7. **Integration**: receive an ERP update through the API and show the change history.
8. **Bulk change**: update a price-relevant attribute across 10,000 products and report the time taken.

Send the scripts and a data sample of 200 to 1,000 products to each vendor two weeks before the demo. Include assets, at least two locales and one deliberately messy file.

## How should a demo day be structured?

The agenda below fits a full day for a mid-market or enterprise selection.

| Time | Session | Who attends |
|---|---|---|
| 09:00 to 09:30 | Introductions and vendor overview (strictly timed) | Whole team |
| 09:30 to 12:30 | Scenarios 1 to 4, with scoring after each | Business users, product data lead |
| 12:30 to 13:15 | Break and individual scoring | Evaluators |
| 13:15 to 15:15 | Scenarios 5 to 8 | Business users, IT, architect |
| 15:15 to 16:00 | Hands-on: two users repeat one scenario unaided | Daily users |
| 16:00 to 16:45 | Technical and security questions | IT, security |
| 16:45 to 17:15 | Implementation approach and commercial outline | Sponsor, procurement |

Schedule the three vendor days within one or two weeks so that impressions stay comparable. Ask the implementation partner who would deliver the project to attend.

## How do you score a scripted demo?

Score each scenario on a 0 to 3 scale with written definitions:

- **0**: not shown or not possible.
- **1**: possible only with custom development or a workaround.
- **2**: shown with configuration, with minor gaps or extra steps.
- **3**: shown fully with standard configuration and usable by a business user.

Multiply each score by the scenario weight agreed before the demos. Add separate scores for usability, partner competence and answers to technical questions. Usability criteria from ISO/IEC 25010:2023, such as learnability and operability, help evaluators judge the hands-on session consistently.

Each evaluator scores alone during the session. Hold a calibration meeting the same afternoon, discuss gaps larger than one point and record the reasons. Fixing weights in advance matches the transparency rule in Article 67(5) of Directive 2014/24/EU and protects private buyers against late bias.

## What rules keep demos comparable?

Set ground rules in the invitation and enforce them on the day:

- Vendors follow the script order and may add extras only after all scenarios are done.
- Slides are limited to the 30-minute opening slot.
- "Roadmap" answers are recorded as a score of 0 for the current release.
- The same evaluators attend all vendor days.
- Recordings are made with the vendor's consent for later review.

The finalist then moves into a short proof of concept, as described in the [eight-step PIM selection process](https://pimatlas.com/knowledge/pim-selection-process/).

## Key takeaways

- Give every vendor the same six to eight scenarios and your own product data two weeks ahead.
- Run one structured day per vendor, including a hands-on session for daily users.
- Score each scenario on a defined 0 to 3 scale, weighted before the demos start.
- Score individually, then calibrate as a team on the same day.
- Treat roadmap answers as not available and verify RFP claims live.

Q: How long should a PIM vendor demo be?
A: Allow one full working day per vendor for a mid-market or enterprise selection, or a half day for a simple SMB catalogue. A full day covers six to eight scenarios with time for hands-on testing, technical questions and a commercial session. Shorter demos tend to skip the hardest scenarios, which are precisely the ones that separate vendors.
Q: Should vendors use our own data in a PIM demo?
A: Yes. Send a sample of 200 to 1,000 real products covering your most complex families, variants, locales and assets, plus a real supplier file. Demos on the vendor's standard data show what the product can do in ideal conditions. Demos on your data show how it handles your attributes, data quality issues and volumes, which is what you will live with after go-live.
Q: How do you score a PIM demo fairly?
A: Use a shared rubric with a 0 to 3 scale per scenario and written definitions for each score. Every evaluator scores independently during the demo, then the team holds a calibration session the same day to discuss differences larger than one point. Weight scenarios in advance and keep the same evaluators for every vendor so the totals remain comparable.

## Article: PIM total cost of ownership over five years
URL: https://pimatlas.com/knowledge/pim-total-cost-of-ownership/
Category: Selection and buying | Updated: 2026-10-17

Summary: PIM total cost of ownership (TCO) is the full five-year cost of a PIM system: subscription or licence, implementation, integrations, data migration, internal staff, enhancements and exit. In the illustrative mid-market model in this article, the subscription is only about a quarter of five-year TCO, while people and implementation make up most of the rest.

PIM total cost of ownership (TCO) is the complete cost of buying, implementing, running and eventually leaving a PIM (product information management) system, measured over five years. It matters because the subscription is usually the smaller part of the spend. Internal staff, implementation partners, integrations and data work typically cost more than the software itself.

## What cost categories make up PIM TCO?

Split the model into one-off and recurring costs so that nothing is counted twice.

**One-off costs**
- Selection: internal time, external advice and any paid proof of concept.
- Implementation partner: configuration, data modelling, testing and project management.
- Integration build: connectors or middleware for ERP, commerce platform, DAM and channels.
- Data migration and cleansing: mapping, conversion and correction of legacy data, as the [data migration guide](https://pimatlas.com/knowledge/pim-data-migration/) explains.
- Internal project team: product data, IT and business staff seconded to the project.
- Initial training and change management.

**Recurring costs**
- Subscription or licence, including annual uplifts and paid modules.
- Hosting, monitoring and security patching if self-hosted.
- Integration maintenance when connected systems change.
- Internal run team: PIM administration, data stewardship and support.
- Enhancements: new channels, attributes, workflows and AI features.
- Exit reserve: export and migration effort at end of life.

## What does a five-year PIM TCO model look like?

The model below is illustrative, built from assumed inputs for a mid-market company with 60,000 SKUs, eight locales, 25 editors and three core integrations. It is not a market benchmark. All figures are in euros.

| Cost line | Year 1 | Year 2 | Year 3 | Year 4 | Year 5 | Five-year total |
|---|---|---|---|---|---|---|
| SaaS subscription (5% annual uplift) | 50,000 | 52,500 | 55,100 | 57,900 | 60,800 | 276,300 |
| Implementation partner | 150,000 | 0 | 0 | 0 | 0 | 150,000 |
| Integrations (build, then maintenance) | 60,000 | 15,000 | 15,000 | 15,000 | 15,000 | 120,000 |
| Data migration and cleansing | 40,000 | 0 | 0 | 0 | 0 | 40,000 |
| Internal team (project, then 1 FTE run) | 115,000 | 70,000 | 70,000 | 70,000 | 70,000 | 395,000 |
| Enhancements and training | 10,000 | 30,000 | 30,000 | 30,000 | 30,000 | 130,000 |
| **Total** | **425,000** | **167,500** | **170,100** | **172,900** | **175,800** | **1,111,300** |

In this illustration the subscription is about 25% of five-year TCO. The internal team is about 36%, and implementation, integrations and data migration together are about 28%. Year one costs roughly 2.5 times any later year.

## How do deployment models change TCO?

Replace the subscription line to compare deployment options. A self-hosted open source edition removes the licence fee but adds hosting, monitoring and upgrade projects. In the model above, assume EUR 10,000 a year for hosting and EUR 25,000 a year from year two for upgrades and security patching. That adds EUR 150,000 against the EUR 276,300 subscription saving, before any extra development.
Commercial on-premise licences add their own lines. Pimcore, for example, publishes Professional and Enterprise Edition prices of USD 9,900 and USD 29,900 per year, while AtroCore sells on-premise premium modules as one-off licences with an optional annual update subscription of 20% of the licence price. The [SaaS, private cloud or on-premise guide](https://pimatlas.com/knowledge/saas-vs-on-premise-pim/) and the [open source PIM guide](https://pimatlas.com/knowledge/open-source-pim/) explain the operational differences.

## Which assumptions move TCO the most?

Test the model against the variables that change the result most:

1. **Renewal uplift**: a 10% annual uplift instead of 5% adds about EUR 29,000 over five years in this model.
2. **SKU and channel growth**: tiered pricing can jump a band when the catalogue or channel count grows.
3. **Data quality at the start**: poor source data can double migration and cleansing effort.
4. **Integration count**: each additional system adds build and maintenance cost every year.
5. **Internal capacity**: under-resourced run teams lead to backlogs and more partner spend.

Contract terms limit several of these. Cap renewal uplifts, fix overage rates and agree exit support. For cloud services, the EU Data Act, Regulation (EU) 2023/2854, abolishes switching charges from 12 January 2027. Internal migration effort still remains part of exit cost.

## How should TCO feed the vendor decision?

Ask each finalist to complete the same five-year pricing template, then add your own internal and partner costs to build one TCO per vendor. Public buyers in the EU can award contracts on a cost-effectiveness basis such as life-cycle costing under Article 67(2) of Directive 2014/24/EU, and private buyers can use the same approach. Compare TCO against the benefits in the [PIM business case](https://pimatlas.com/knowledge/pim-business-case/) and against published price points in the [PIM pricing guide](https://pimatlas.com/knowledge/pim-pricing-models/).

## Key takeaways

- PIM TCO covers subscription, implementation, integrations, data migration, internal staff, enhancements and exit over five years.
- In the illustrative mid-market model, the subscription is about 25% of TCO and internal staff about 36%.
- Self-hosting removes licence fees but adds hosting, upgrades and development that narrow the saving.
- Renewal uplifts, catalogue growth, data quality and integration count move TCO the most.
- Build one TCO per finalist from a common pricing template and your own internal costs.

Q: What is included in PIM total cost of ownership?
A: PIM total cost of ownership includes one-off costs (selection, implementation partner, integration build, data migration, internal project team, training) and recurring costs (subscription or licence, hosting if self-hosted, integration maintenance, internal run team, enhancements, support and renewal uplifts). It should also include the expected cost of exit, such as data export and migration to a successor system.
Q: Why compare PIM vendors over five years instead of one?
A: Year-one costs are dominated by implementation, while years two to five are dominated by subscription, uplifts and internal staff. A vendor with a low first-year price and uncapped renewals can cost more over five years than a vendor with a higher entry price. Five years also matches the typical life of a PIM platform before major re-platforming.
Q: Is the PIM subscription the largest part of TCO?
A: Usually not. In the illustrative mid-market model in this article, the subscription accounts for about 25% of five-year TCO, while internal staff account for about 36% and implementation, integrations and data migration for about 28%. The exact split varies by organisation, so build your own model with real quotes and internal salary costs.

## Article: A PIM implementation roadmap: phases, timelines and team
URL: https://pimatlas.com/knowledge/pim-implementation-roadmap/
Category: Implementation and data | Updated: 2026-10-17

Summary: A PIM implementation typically runs through five phases: discovery, data model design, configuration and integration, data migration, and go-live with stabilisation. Vendor benchmarks put a focused mid-market rollout at roughly 3 to 6 months and a multi-region enterprise programme at 6 to 18 months. Data readiness and decision speed drive the timeline more than SKU count.

A successful PIM (product information management) implementation follows a phased roadmap. It starts with discovery and data model design, moves through configuration, integration and migration, and finishes with a staged go-live. Published vendor benchmarks put a focused rollout at about three to six months and a complex enterprise programme at six to eighteen months. Most delays trace back to data readiness and slow decisions, and software configuration accounts for very few of them.

## What are the phases of a PIM implementation?

Most methodologies share the same five phases, whatever the vendor. The durations below are typical ranges taken from published benchmarks by Inriver and Plytix. They are planning guides only, and a given project will differ.

| Phase | Main outputs | Typical mid-market | Typical enterprise |
|---|---|---|---|
| 1. Discovery and scoping | Use cases, channel list, source system inventory, success KPIs | 2 to 4 weeks | 4 to 8 weeks |
| 2. Data model design | Families, attributes, variants, taxonomy, validation rules | 2 to 4 weeks | 6 to 10 weeks |
| 3. Configuration and integration | Workflows, roles, ERP, DAM and commerce connectors | 4 to 6 weeks | 6 to 12 weeks |
| 4. Data migration | Cleansed, mapped and loaded product records | 4 to 8 weeks | 8 to 16 weeks |
| 5. Go-live and stabilisation | Cut-over, hypercare, first channel live | 2 to 4 weeks | 4 to 6 weeks |

Phases overlap in practice. Migration scripting can start once the data model is stable, and integration work often runs alongside configuration. The overall elapsed time is therefore shorter than the sum of the columns.

## Why do discovery and data modelling decide the timeline?

Discovery fixes scope. It answers which channels the PIM feeds first, which systems own which attributes, and which KPIs define success (for example, time to market for a new SKU, or the share of products complete for a channel). Teams that skip this step usually reopen scope during configuration, which is the most expensive time to do it.

The data model sets the structure for everything that follows. Families, attribute types, variant axes and the classification scheme determine how imports, workflows and exports behave. Our guide to [product data modelling](https://pimatlas.com/knowledge/product-data-modelling/) covers the design choices in detail. A model signed off by category owners before configuration begins is the single strongest predictor of an on-time project.

## Who should be on the PIM project team?

A PIM project changes how several departments work, so the team spans business and IT. A typical core team includes:

- **Business product owner:** owns scope, priorities and acceptance. Usually from e-commerce, marketing or product management.
- **Project manager:** runs the plan, the budget and the vendor or partner relationship.
- **PIM solution architect or data architect:** designs the data model, the validation rules and the integration contracts.
- **Integration developers:** build and test connectors to ERP, DAM, commerce and syndication tools.
- **Data stewards:** category specialists who enrich data, test workflows and sign off migrated records.
- **Change and training lead:** prepares users, documentation and the support model.

The ownership rules agreed here feed directly into [product data governance](https://pimatlas.com/knowledge/product-data-governance/), so governance roles are best named during the project.

## How should go-live be staged?

A wave-based rollout lowers risk. The first wave usually covers one channel, one region or one product range, which proves the model and integrations under real conditions. Later waves reuse that configuration, so they tend to be shorter.

Plan the cut-over in detail. That means a content freeze in legacy systems, a final delta migration, smoke tests on every outbound feed, and a hypercare period of typically two to six weeks with daily triage. Our article on [PIM data migration](https://pimatlas.com/knowledge/pim-data-migration/) lists the steps for the load itself, and the [integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/) guide explains how to sequence connectors.

## Which risks most often delay a PIM project?

Delays rarely come from the software. Benchmarks from Inriver and Plytix both point to data quality and decision-making as bigger drivers than catalogue size. Common risks include:

1. Source data that needs more cleansing than discovery estimated.
2. Unresolved attribute ownership between ERP, marketing and category teams.
3. Integration dependencies on other projects, such as a parallel ERP or commerce replatforming.
4. Too many custom requirements in the first wave.
5. Data stewards who are not freed from their day jobs for testing and enrichment.

Each risk has a matching mitigation: a data quality assessment in discovery, a signed attribute ownership matrix, a shared integration calendar, a strict first-wave scope, and budgeted steward time. These choices also feed the [PIM selection process](https://pimatlas.com/knowledge/pim-selection-process/), since implementation effort varies by platform and partner.

## Key takeaways

- A PIM roadmap typically has five phases: discovery, data model design, configuration and integration, data migration, and go-live with stabilisation.
- Published benchmarks suggest about 3 to 6 months for a focused rollout and 6 to 18 months for a complex enterprise programme.
- A signed-off data model and an agreed attribute ownership matrix protect the timeline more than any other activity.
- A wave-based go-live, starting with one channel or region, reduces risk and speeds up later waves.
- Budget data steward time explicitly, because enrichment and testing compete with daily work.

Q: How long does a PIM implementation take?
A: Vendor benchmarks give a wide range. Inriver cites three to twelve months overall, and Plytix cites four to twelve weeks for a small or mid-market minimum viable rollout and six to eighteen months for a complex enterprise programme. Catalogue complexity, the number of integrations, data quality and how quickly the business makes decisions move a project within those ranges.
Q: Who should be on a PIM implementation team?
A: A typical core team has a business product owner, a project manager, a data architect or PIM solution architect, one or more integration developers, data stewards from the category teams and a change lead. The vendor or an implementation partner usually supplies configuration specialists. Executive sponsorship matters most when attribute ownership has to move between departments.
Q: Should a PIM go live with all channels at once?
A: Most programmes reduce risk by going live with one channel, one region or one product range first, then adding the rest in waves. A first wave proves the data model, the integrations and the workflow under real load. Later waves reuse that configuration, so each one is usually shorter than the first.

## Article: Product data modelling: families, attributes and variants
URL: https://pimatlas.com/knowledge/product-data-modelling/
Category: Implementation and data | Updated: 2026-10-17

Summary: A PIM data model groups products into families that share an attribute set, defines each attribute with a type, scope and validation rule, and organises variants as a parent with child products that differ along defined axes such as size or colour. A good model mirrors how customers choose products and how channels require data.

A product data model defines how a PIM (product information management) system stores product information. It rests on three building blocks: families, attributes and variants. Families group products that share an attribute set. Attributes define each piece of data with a type, a scope and validation rules. Variants organise sellable items under a parent along axes such as colour or size. When the model follows how customers choose products, enrichment, quality scoring and channel exports all become simpler.

## What are families, attributes and variants?

A **family** is a set of products described by the same attributes, such as "cordless drills" or "dining chairs". The family decides which attributes appear and which are required for each channel.

An **attribute** is one property of a product, for example voltage, material or marketing description. Each attribute has a type (text, number with unit, select list, boolean, date, asset, reference), and it may be localisable, channel-specific or both.

A **variant** is a sellable item that shares most content with siblings and differs along one or more **variant axes**. In Akeneo's documented model, a "family variant" distributes a family's attributes across levels, a "product model" holds the shared values, and variant products hold the values that differ. Other platforms use different terms, but the logic is the same.

## What does a worked data model look like?

The table below shows an illustrative model for three families. The attribute names and splits are examples only.

| Family | Shared attributes (parent) | Variant axes | Variant-level attributes |
|---|---|---|---|
| T-shirts | Brand, fit, fabric composition, care instructions, description | Level 1: colour; level 2: size | Colour images (level 1); GTIN, SKU, stock unit (level 2) |
| Cordless drills | Brand, chuck size (mm), max torque (Nm), battery platform | Kit contents | GTIN, SKU, number of batteries, case included |
| LED panels | Dimensions (mm), IP rating, mounting type | Colour temperature (K) | Luminous flux (lm), GTIN, ETIM class EC number |

The rule of thumb: put an attribute at the highest level where it is always identical. Images often vary by colour and almost never by size, so they belong at colour level. Identifiers such as the GTIN (Global Trade Item Number) always sit at the sellable level.

## How should attribute types and rules be designed?

Strong typing is the cheapest data quality control available. A free-text "voltage" field collects "18V", "18 V" and "18 volt". A numeric attribute with a unit stores one comparable value.

Good practice for attribute design:

- **Use select lists** for any value set that can be enumerated, such as material, colour family or certification.
- **Store numbers with units** and convert on export, never in the source value.
- **Separate marketing text from technical data**, so translation and approval workflows can differ.
- **Decide scope explicitly**: global, per locale, per channel, or per locale and channel.
- **Add validation** (regular expressions, minimum and maximum values, allowed file types) at attribute level.
- **Name attributes consistently**, with a code that never changes and a label that can be translated.

B2B standards follow the same principle. In the ETIM model, every feature is alphanumeric (a fixed value list), logical, numeric or range, and numeric features carry a unit. ETIM 11.0 is planned for final release on 1 December 2026, so teams aligning with ETIM should plan a mapping update. Our guide to [ETIM, BMEcat and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/) explains the standards.

## How do families relate to the product taxonomy?

Families describe data structure, while categories describe navigation and classification. A single family such as "chairs" may appear in several web categories ("office", "dining", "outdoor"). Keeping the two separate lets merchandisers reshape navigation without restructuring data. The companion article on [product taxonomy design](https://pimatlas.com/knowledge/product-taxonomy-design/) covers category trees and standard classifications.

## Which data modelling mistakes are hardest to undo?

Some decisions become expensive once products and integrations depend on them:

1. **One giant family** with hundreds of optional attributes, which breaks completeness scoring.
2. **Variant axes chosen for convenience**, such as using "pack size" as an axis when packs have different GTINs and content.
3. **Unit and language baked into values**, which blocks conversion and translation.
4. **Attribute codes that mirror one system**, such as ERP field names, which tie the model to that system's limits.
5. **No reference entities** for brands, suppliers or materials, which leads to duplicated text.

Review the model with category owners and channel managers before migration starts, as described in the [PIM implementation roadmap](https://pimatlas.com/knowledge/pim-implementation-roadmap/). The model also sets the baseline for [product data quality](https://pimatlas.com/knowledge/product-data-quality/) metrics.

## Key takeaways

- A PIM data model rests on families, typed attributes and variants organised along explicit axes.
- Place each attribute at the highest level where its value is always identical, and keep identifiers at the sellable level.
- Strong typing, units and select lists prevent most quality issues before they occur.
- Keep families (data structure) separate from categories (navigation and classification).
- Align with ETIM or ECLASS early where B2B partners require them, and plan for the ETIM 11.0 release.

Q: What is a product family in a PIM?
A: A product family is a group of products that share the same set of attributes, such as all drills or all T-shirts. The family defines which attributes appear, which are required for each channel and how completeness is calculated. Families are a data-structure concept and usually differ from the navigation categories shown to shoppers on a website.
Q: How many variant levels should a product model have?
A: Most catalogues need one or two variant levels. A T-shirt might use colour at the first level and size at the second, so that images and descriptions sit at colour level and stock-keeping data sits at size level. Additional levels add maintenance effort, so each level should match a real difference in content or channel requirements.
Q: Should a PIM data model follow ETIM or ECLASS?
A: Where trading partners require ETIM or ECLASS, align families and attributes to those classes from the start, or map to them in a dedicated classification. ETIM features have defined types and units, which makes mapping straightforward. Companies that sell only through consumer channels can design their own model and map to standards where channels need them.

## Article: Designing a product taxonomy and classification scheme
URL: https://pimatlas.com/knowledge/product-taxonomy-design/
Category: Implementation and data | Updated: 2026-10-17

Summary: A sound product taxonomy separates three layers: an internal master classification that drives attributes, channel-specific navigation trees for shoppers, and mappings to external standards such as GS1 GPC, ETIM, ECLASS and Google's product taxonomy. Each layer has its own owner and change process, and the PIM maintains the links between them.

A product taxonomy works best as three linked layers. The first is an internal master classification that decides which attributes each product needs. The second is a set of navigation trees built for shoppers on each channel. The third is a set of mappings to external standards such as GS1 GPC, ETIM, ECLASS and the Google product taxonomy. A PIM (product information management) system holds all three layers and the links between them, which lets each layer change without breaking the others.

## What is a product taxonomy?

A product taxonomy is a hierarchy that classifies products into categories, from broad to specific, such as Lighting > Indoor lighting > LED panels. A classification scheme is the formal version of that idea: a controlled set of classes, often with codes, definitions and attributes attached to each class.

In a PIM, taxonomy serves three purposes. It determines the attribute set through the link to [product families](https://pimatlas.com/knowledge/product-data-modelling/). It drives navigation and filtering on websites and in print catalogues. It also supplies the class codes that retailers, wholesalers and marketplaces require.

## Which external classification standards matter?

Each standard serves a different trading context, and their structures differ. The facts below come from the standards bodies' own documentation.

| Standard | Typical use | Structure | Notes |
|---|---|---|---|
| GS1 GPC (Global Product Classification) | Consumer goods, GDSN data synchronisation | Segment > family > class > brick | Published twice a year and date-stamped rather than versioned. A placeholder brick of 99999999 is used where no brick fits yet. |
| ETIM | Electrical, HVAC, plumbing, tools, building products | Product group > product class, with typed features | Classes have language-independent IDs and local-language synonyms. ETIM 11.0 is planned for 1 December 2026. |
| ECLASS | Industrial procurement and engineering | Segment > main group > group > commodity class | Eight-digit codes, two digits per level. Properties attach only at level four. |
| Google product taxonomy | Google Shopping and Merchant Center feeds | Multi-level category paths | Optional attribute. Google auto-assigns a category, and merchants can submit an ID or a full path to override it. |

Our guides to [ETIM, BMEcat and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/) and to [GS1, GTINs and GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/) cover each standard in more depth.

## How should the master classification be designed?

The master classification is the internal backbone. Its job is to assign the right attribute set and the right owner to every product. Design criteria that hold up in practice:

- **Mutually exclusive leaves:** every product belongs to exactly one master class.
- **Attribute-driven splits:** create a new leaf only when the products need a different attribute set or a different owner.
- **Stable codes:** class codes never change. Labels can be renamed and translated.
- **Definitions and examples:** each class has a written scope note with included and excluded examples.
- **Depth guided by data:** typically three to five levels, deeper only where the attributes justify it.
- **External mappings:** map each leaf to GPC, ETIM, ECLASS or Google classes. Adopt one external standard as the master only when a single trading context dominates.

## How do navigation trees differ from the master classification?

Navigation trees serve shoppers. They group products by intent ("garden lighting", "gifts under 50 euros"), may repeat products across branches, and change with seasons and campaigns. Typical shopper-facing trees stay at three or four levels.

Keep navigation separate from the master classification in the PIM, or in the commerce platform, with rule-based assignment where possible. A rule such as "family = LED panel and IP rating ≥ 44 → Outdoor lighting" keeps the tree current without manual tagging.

## How should taxonomy changes be governed?

Taxonomies drift without a process. A practical change process includes:

1. A named taxonomy owner, usually within the product data or merchandising team.
2. A change request log with rationale, affected products and affected channels.
3. Impact analysis on attributes, completeness scores and outbound mappings before approval.
4. Scheduled updates aligned with external releases, such as the twice-yearly GPC publications and ETIM major versions.
5. Regression checks on feeds after each change.

These steps belong in the wider [product data governance](https://pimatlas.com/knowledge/product-data-governance/) model, and they protect [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/) feeds from silent breakage.

## Key takeaways

- Treat taxonomy as three layers: master classification, channel navigation trees and external standard mappings.
- Split master classes only when the attribute set or the owner changes, and keep class codes stable.
- GS1 GPC uses segment, family, class and brick. ECLASS uses four coded levels. ETIM uses groups and classes with typed features.
- Navigation trees should follow shopper intent and can be rule-based on attribute values.
- Govern taxonomy changes with a named owner, impact analysis and a calendar aligned to GPC and ETIM releases.

Q: What is the difference between a taxonomy and a product family?
A: A taxonomy is a hierarchy of categories used to classify and find products, such as Tools > Power tools > Drills. A product family is a data structure that defines which attributes a product has. One family can appear in several taxonomy branches, and one category can contain several families, so the two are usually maintained separately in a PIM.
Q: Which classification standard should a manufacturer use?
A: It depends on the trading partners. Consumer goods suppliers that synchronise data through GDSN need a GS1 GPC brick for each item. Electrical, installation and building products wholesalers in Europe commonly require ETIM. Industrial procurement platforms often use ECLASS. Many manufacturers maintain mappings to two or three standards from one internal classification.
Q: How deep should a product category tree be?
A: There is no single correct depth. Shopper-facing trees typically work best with three or four levels, so that most products are reachable in a few clicks. Internal master classifications can go deeper when each leaf genuinely needs a different attribute set. Test any tree with real search and navigation data before finalising it.

## Article: Product data quality: metrics, completeness and scoring
URL: https://pimatlas.com/knowledge/product-data-quality/
Category: Implementation and data | Updated: 2026-10-17

Summary: Product data quality is measured across six dimensions: completeness, validity, accuracy, consistency, uniqueness and timeliness. Each needs a formula, a target and an owner. A PIM calculates completeness per family, channel and locale. A weighted score that combines several dimensions shows which products are ready to publish and where to direct enrichment effort.

Product data quality is managed through measurement. Teams track six dimensions (completeness, validity, accuracy, consistency, uniqueness and timeliness), and each needs a formula, a target and an owner. A PIM (product information management) system calculates completeness per family, channel and locale out of the box. A weighted score that combines several dimensions then shows which products are ready to publish and where enrichment effort will pay back first.

## What are the dimensions of product data quality?

The UK Government Data Quality Hub, drawing on DAMA UK, names six dimensions. Each one maps to a concrete product data problem:

- **Completeness:** required attributes have values, such as a missing net weight blocking a retailer listing.
- **Validity:** values conform to type, format and range, such as a GTIN with a wrong check digit.
- **Accuracy:** values reflect the physical product, such as declared dimensions matching the measured pack.
- **Consistency:** values agree across attributes and systems, such as the colour in the title matching the colour attribute.
- **Uniqueness:** each product exists once, with no duplicate SKUs or GTINs.
- **Timeliness:** data is available when the channel needs it, such as before a launch date.

## How do you calculate data quality metrics?

Each dimension needs a repeatable formula. The table shows common definitions with illustrative targets. Calibrate the targets to your own channels.

| Metric | Formula | Illustrative target |
|---|---|---|
| Completeness (per channel and locale) | Required attributes filled ÷ required attributes × 100 | 100% before publish |
| Validity rate | Values passing validation rules ÷ values checked × 100 | 99% or higher |
| Accuracy rate | Attributes matching a physical or source audit ÷ attributes audited × 100 | 98% or higher on audited samples |
| Consistency rate | Records with no cross-system mismatches ÷ records compared × 100 | 98% or higher |
| Duplicate rate | Duplicate records ÷ total records × 100 | Below 0.5% |
| Time to ready | Date channel-complete minus date created, in days | Agreed per category |
| Enrichment backlog | Products below threshold for a priority channel | Downward trend |

Completeness is usually native to the PIM. In Akeneo, for example, completeness depends on the family, the channel and the locale, and it is recalculated after imports, rules and saves. Accuracy is the hardest to automate, because it requires comparison with the real product. The GS1 Data Quality Framework includes a Data Inspection Procedure for physically validating attributes.

## How should a weighted quality score work?

A single score helps teams prioritise, provided it stays transparent. A common illustrative model:

1. Set channel completeness as a gate. Products below 100 per cent on mandatory attributes score zero for that channel.
2. Weight the remaining dimensions, for example validity 30 per cent, enrichment completeness 30 per cent, consistency 20 per cent and asset coverage 20 per cent.
3. Score by family and channel, because a fashion marketplace and a B2B portal value different attributes.
4. Publish the score next to its components, so stewards can see what to fix.

Scores work best when they trigger action: workflow tasks, dashboard alerts or blocked exports.

## Where should quality rules be enforced?

Prevention costs less than correction. Rules belong at four points:

- **Data model:** typed attributes, units, select lists and required flags, as covered in [product data modelling](https://pimatlas.com/knowledge/product-data-modelling/).
- **Input:** validation on supplier imports and onboarding portals.
- **Workflow:** approval steps that block publication until checks pass.
- **Output:** channel-specific checks before syndication, such as retailer or GDSN validation rules described in our guide to [GS1, GTINs and GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/).

[AI enrichment](https://pimatlas.com/knowledge/ai-product-data-enrichment/) can fill gaps at scale, but generated values need the same validation and a human review sample.

## Which standards support product data quality?

Three references are widely used. The GS1 Data Quality Framework contains a data quality management system, self-assessment tools and the inspection procedure. ISO 8000 is a multi-part international data quality standard. ISO 8000-110:2021 covers the exchange of master data as characteristic data, and separate parts address provenance (ISO 8000-120), accuracy (ISO 8000-130) and completeness (ISO 8000-140). DAMA's body of knowledge frames quality within wider data management.

Metrics only hold when someone owns them. The roles that do so are set out in [product data governance](https://pimatlas.com/knowledge/product-data-governance/).

## Key takeaways

- Measure six dimensions: completeness, validity, accuracy, consistency, uniqueness and timeliness.
- Give every metric a formula, a target, a reporting frequency and an owner.
- Use channel completeness as a publication gate, then weight the other dimensions into a transparent score.
- Enforce rules in the data model, at input, in workflow and at output, and use reports to confirm they work.
- The GS1 Data Quality Framework and ISO 8000 provide external reference points for audits.

Q: How is product completeness calculated in a PIM?
A: Completeness is usually the number of required attributes that have a value, divided by the total number of required attributes, for a given product, channel and locale. Required attributes are defined per family and per channel, so the same product can be 100 per cent complete for a web shop and 70 per cent complete for a marketplace with stricter requirements.
Q: What is a good product data quality score?
A: Targets depend on the channel and the attribute. Mandatory attributes for a channel should be 100 per cent complete before publishing, because many retailers and marketplaces reject incomplete records. Organisations typically set lower targets for enrichment attributes and raise them over time. Report trends weekly, since a single snapshot hides regressions.
Q: Which standards define data quality for product data?
A: The GS1 Data Quality Framework provides a data quality management system, self-assessment tools and a physical data inspection procedure for trade item data. ISO 8000 is a multi-part international standard for data quality, with Part 110 covering the exchange of master data. The DAMA data management community describes the common quality dimensions.

## Article: Product data governance: roles, workflows and ownership
URL: https://pimatlas.com/knowledge/product-data-governance/
Category: Implementation and data | Updated: 2026-10-17

Summary: Product data governance assigns clear ownership for every attribute, defines who creates, enriches, approves and publishes product data, and sets the policies and workflows a PIM enforces. A typical model has a governance council, data owners, data stewards and technical custodians, documented in a RACI matrix and backed by measurable quality targets.

Product data governance is the set of roles, rules and workflows that decides who may create, change, approve and publish each piece of product information. It works when every attribute group has a named owner, every change follows a defined workflow and every policy links to a measurable quality target. A PIM (product information management) system enforces governance through permissions, workflows and audit trails, but people and policies have to come first.

## What does product data governance cover?

Governance answers four questions for every attribute: who is accountable, who maintains it, which rules it must meet, and how changes are approved. In practice that covers:

- **Ownership:** a named owner per attribute group and per product category.
- **Definitions:** a business glossary with attribute meaning, format, unit and allowed values.
- **Policies:** rules for identifiers, naming, imagery, translations and regulatory content.
- **Workflows:** creation, enrichment, review, approval and retirement steps.
- **Measurement:** the quality metrics described in [product data quality](https://pimatlas.com/knowledge/product-data-quality/).
- **Change control:** how the data model and the [product taxonomy](https://pimatlas.com/knowledge/product-taxonomy-design/) change over time.

DAMA International's DMBOK (Data Management Body of Knowledge) treats data governance as the function that guides all other data management activities. That framing fits product data well.

## Which roles does a governance model need?

Most product data governance models use four layers of responsibility:

1. **Governance council or steering group:** senior representatives from e-commerce, product management, supply chain, marketing, compliance and IT. Sets priorities and resolves ownership disputes.
2. **Data owners:** accountable managers for a data domain, such as technical data, marketing content or logistics data. Approve definitions and targets.
3. **Data stewards:** specialists who create, enrich and correct records, triage quality issues and propose rule changes.
4. **Data custodians:** IT or platform teams that run the PIM, integrations, security and backups.

Smaller organisations often combine the council and owner roles. The principle stays the same: accountability for each attribute sits with one named person.

## What does a RACI matrix for product data look like?

A RACI matrix (responsible, accountable, consulted, informed) turns ownership into a working agreement. The example below is illustrative. Adjust the departments to your organisation.

| Activity | Product mgmt | Marketing / e-commerce | Supply chain | Compliance | PIM team (IT) |
|---|---|---|---|---|---|
| Create new SKU and identifiers | A | I | R | I | C |
| Technical attributes and specifications | A/R | C | I | C | I |
| Marketing copy and imagery | C | A/R | I | I | I |
| Logistics data (dimensions, weights) | I | I | A/R | I | I |
| Safety and regulatory information | C | I | I | A/R | I |
| Publish to channels | I | A | I | C | R |
| Change the data model or taxonomy | C | C | C | C | R, with council as A |

Identifier rules deserve an explicit policy. The GS1 GTIN Management Standard says a new GTIN is required only when consumers or trading partners need to distinguish the product, when a regulatory or liability disclosure applies, or when the change materially affects the supply chain. Our guide to [GS1, GTINs and GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/) explains the rules.

## How do workflows enforce governance in a PIM?

Workflows turn policy into daily practice. A typical new-product workflow runs:

1. **Creation:** supply chain or ERP creates the SKU with core identifiers.
2. **Technical enrichment:** product management completes specifications.
3. **Marketing enrichment:** copywriters and the DAM team add content and assets.
4. **Compliance review:** regulated attributes are checked and approved.
5. **Quality gate:** the PIM checks channel completeness and validation rules.
6. **Publication:** approved products flow to channels.

Each step needs a service level, such as an illustrative three working days for marketing enrichment, so that bottlenecks show up in reporting. Permissions should restrict edit rights by attribute group, and audit history should record every change.

## How do you introduce governance without slowing the business?

Start small and tie governance to an outcome the business already wants, such as faster launches or fewer marketplace rejections. A practical sequence:

- Name owners for the attribute groups that cause the most errors.
- Publish a short glossary for the top 50 to 100 attributes.
- Automate the first quality gate before adding approval steps.
- Review metrics monthly in the governance council.

Governance design fits best in the discovery phase of the [PIM implementation roadmap](https://pimatlas.com/knowledge/pim-implementation-roadmap/), when ownership questions surface anyway. Companies with wider master data needs should also read [PIM vs MDM](https://pimatlas.com/knowledge/pim-vs-mdm/).

## Key takeaways

- Governance assigns a named owner, rules and a workflow to every product attribute group.
- Use four layers: a governance council, data owners, data stewards and technical custodians.
- A RACI matrix per activity makes ownership explicit and settles disputes before they reach production.
- Enforce governance in the PIM through permissions, workflows, quality gates and audit history.
- Start with the attributes that cause the most errors and link governance to a measurable business outcome.

Q: What is the difference between a data owner and a data steward?
A: A data owner is accountable for a data domain, such as technical specifications or marketing content, and approves policies, definitions and quality targets for it. A data steward does the day-to-day work: enriching, correcting and validating records against those policies. Owners are usually senior business managers. Stewards are usually category or content specialists.
Q: Who should own product data in a company?
A: Ownership is usually split by attribute group. Product management or engineering typically owns technical attributes, marketing or e-commerce owns descriptive content and imagery, supply chain owns logistics data such as dimensions and weights, and regulatory or quality teams own compliance data. The PIM then enforces who can edit and approve each group.
Q: Does product data governance require a separate tool?
A: Most PIM systems include the core governance controls: role-based permissions, workflows, approval steps, audit history and quality dashboards. Larger organisations sometimes add a data catalogue or MDM governance module to manage definitions across domains. The policies, roles and change processes matter more than the tool chosen.

## Article: Migrating product data into a new PIM
URL: https://pimatlas.com/knowledge/pim-data-migration/
Category: Implementation and data | Updated: 2026-10-17

Summary: A PIM data migration moves product records, assets and relationships from spreadsheets, ERP, legacy PIM and e-commerce systems into a new data model. It runs in eight steps, from source inventory and profiling to mapping, cleansing, test loads, reconciliation and cut-over. Vendor benchmarks show migration is often the longest phase of a PIM project.

Migrating product data into a new PIM (product information management) system means extracting records, assets and relationships from every current source. That data is then profiled, mapped to the new data model, cleansed, loaded in rehearsals and reconciled before a controlled cut-over. Published vendor benchmarks show migration is often the longest single phase of a PIM project, so it needs its own plan, owner and acceptance criteria.

## Why is PIM data migration the riskiest phase?

Migration exposes every inconsistency that legacy processes have hidden. Spreadsheets hold free-text values, ERP fields have been repurposed, and the same attribute exists in three systems with three different meanings. Inriver's published benchmark places migration at four to eight weeks for smaller businesses and eight to sixteen weeks for enterprises, longer than configuration.

The new PIM also has a stricter [data model](https://pimatlas.com/knowledge/product-data-modelling/) than most sources. Typed attributes, select lists and variant structures reject data that a spreadsheet accepted without complaint.

## What are the steps of a PIM data migration?

The eight steps below form a common sequence. Durations vary widely by project and are shown as illustrative ranges for a mid-sized catalogue.

| Step | What happens | Output | Illustrative effort |
|---|---|---|---|
| 1. Source inventory | List every source: ERP, spreadsheets, legacy PIM, web shop, DAM, supplier files | Source register with owners | 1 week |
| 2. Scope decision | Agree which products, locales and history to migrate | Signed scope | Days |
| 3. Profiling | Measure fill rates, formats, duplicates and outliers per field | Profiling report | 1 to 2 weeks |
| 4. Mapping | Map each source field to a target attribute, with conversion rules | Mapping specification | 2 to 4 weeks |
| 5. Cleansing and conversion | Standardise units, split combined fields, resolve duplicates, map classifications | Cleansed staging data | 2 to 6 weeks |
| 6. Test loads | Load into a test environment, iterate on errors | Load logs, error rate trend | 2 to 4 rehearsals |
| 7. Reconciliation | Compare counts, values and relationships with sources, then steward sign-off | Reconciliation report | 1 to 2 weeks |
| 8. Cut-over | Content freeze, final delta load, switch integrations, hypercare | Live PIM | Days plus hypercare |

## How should source data be mapped?

Mapping is where business knowledge enters the migration. Each target attribute needs a defined source, a conversion rule and a fallback. Common conversion rules include:

- **Unit normalisation:** "18V", "18 V" and "18 volt" become 18 with unit V.
- **Field splitting:** a combined "Colour/Size" column becomes two variant axes.
- **Value mapping:** legacy material codes map to new select-list options.
- **Classification mapping:** legacy categories map to the new master classification and to external codes. For GS1 GPC, the placeholder brick 99999999 is available where no brick fits yet.
- **Asset linking:** file names or URLs resolve to DAM asset IDs. See [DAM and PIM integration](https://pimatlas.com/knowledge/dam-pim-integration/).
- **Source priority:** when ERP and a spreadsheet disagree, a documented rule decides which value wins.

Keep the mapping specification under version control. It becomes the reference for later integrations.

## How do you validate a migration?

Validation needs numbers as well as spot checks. Typical acceptance criteria:

1. Record counts per family match the agreed scope, with every exclusion explained.
2. Mandatory attributes reach an agreed completeness level, measured with the metrics in [product data quality](https://pimatlas.com/knowledge/product-data-quality/).
3. Zero duplicate identifiers (SKU and GTIN).
4. Every variant has a valid parent, and every asset link resolves.
5. Category owners sign off a sample of products per family.

Run the same checks after every test load and plot the error rate. A falling trend over two or three rehearsals is a good signal that the team is ready for cut-over.

## How should the cut-over be planned?

Cut-over planning prevents data loss between the last rehearsal and go-live. Agree a content freeze window in legacy systems, run a final delta load of changes since the last full load, and switch inbound and outbound integrations in a fixed order. The [integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/) guide covers connector sequencing.

Keep the legacy sources read-only for a defined period, typically several weeks, so teams can check historical values. The full context of phases and team roles appears in the [PIM implementation roadmap](https://pimatlas.com/knowledge/pim-implementation-roadmap/).

## Key takeaways

- Migration is often the longest phase of a PIM project. Inriver benchmarks it at 4 to 16 weeks depending on company size.
- Follow eight steps: inventory, scope, profiling, mapping, cleansing, test loads, reconciliation and cut-over.
- Cleanse during migration with documented conversion rules, and keep the mapping specification under version control.
- Validate with numeric acceptance criteria and track the error rate across rehearsals.
- Reduce scope by archiving inactive products instead of migrating everything.

Q: How long does a PIM data migration take?
A: Inriver's published benchmark puts data migration at four to eight weeks for smaller businesses and eight to sixteen weeks for enterprises, which makes it the longest single phase in its implementation table. The main drivers are the number of source systems, the amount of cleansing required and how quickly category owners can review mapped data.
Q: Should product data be cleansed before or after migration?
A: Most cleansing is best done during migration, between extraction and loading, where scripts and rules can be applied consistently. Fixing data in legacy systems first is often slow, and loading dirty data into the new PIM spreads errors to channels. Leave only low-priority enrichment for after go-live, and track it as a backlog.
Q: Should all historical products be migrated into the PIM?
A: Usually not. Migrate active and recently discontinued products, plus anything still needed for spare parts, warranty or regulatory retention. Archive the rest in a read-only store. Reducing scope early shortens mapping, cleansing and testing, and keeps the new PIM focused on products that channels actually need.

## Article: PIM integration architecture: ERP, commerce, DAM and APIs
URL: https://pimatlas.com/knowledge/pim-integration-architecture/
Category: Implementation and data | Updated: 2026-10-17

Summary: A PIM integration architecture sets which system owns each data domain and how data moves between them. ERP usually owns identifiers, prices and logistics, DAM owns assets, and the PIM owns enriched product content. Data moves by scheduled batch, request-driven APIs or event-driven messaging, often through an integration platform, and MACH principles shape composable stacks.

A PIM (product information management) integration architecture rests on two decisions: which system is the record holder for each data domain, and which pattern moves data between systems. ERP usually owns identifiers, prices and logistics data. DAM owns master assets. The PIM owns enriched, channel-ready content. Data moves by scheduled batch, request-driven APIs or event-driven messaging, and larger landscapes route these flows through an integration platform.

## Which system should own which product data?

Clear ownership prevents two systems from overwriting each other. A typical split, which every organisation should confirm in its own [PIM and ERP](https://pimatlas.com/knowledge/pim-vs-erp/) design:

| Data domain | Typical system of record | Direction relative to PIM |
|---|---|---|
| SKU creation, core identifiers, GTIN | ERP | Inbound |
| Prices, stock, lead times | ERP or pricing engine | Usually bypasses PIM, straight to commerce |
| Engineering specifications, BOM | PLM | Inbound |
| Master images, videos, documents | DAM | Inbound as links or renditions |
| Marketing copy, enriched attributes, translations | PIM | Outbound |
| Channel feeds for retailers and marketplaces | PIM or syndication tool | Outbound |
| Web navigation, merchandising | Commerce platform | Inbound category assignments or outbound |

Fast-changing data such as stock and price rarely belongs in the PIM. Routing it through the PIM adds latency and load without adding value.

## Which integration patterns are used with PIM?

Four patterns cover most needs. Real landscapes mix them.

- **Scheduled batch (file or bulk API):** full or delta exports on a timetable, such as a nightly CSV or XML file to a distributor. Simple and predictable, but data is only as fresh as the last run.
- **Request-driven API:** a system calls the PIM's REST or GraphQL API when it needs data. Pimcore's Datahub, for example, exposes configurable GraphQL endpoints with queries and mutations. Suited to headless front ends and on-demand lookups.
- **Event-driven messaging:** the PIM publishes events when products change, and subscribers react. Akeneo's Event Platform sends events such as product updated or product became complete, in CloudEvents format, to Google Pub/Sub, HTTPS, Kafka or AMQP destinations. Near real-time, at the cost of designing for retries and ordering.
- **Integration platform (iPaaS or ESB):** a central layer handles mapping, routing, monitoring and retries for all of the above. It adds a component, but it removes point-to-point sprawl.

Event-driven designs need safeguards. Akeneo's documentation, for example, notes that HTTPS endpoints must respond within five seconds and that events are lost while a subscription is suspended. A periodic reconciliation batch is a sensible backstop.

## How do MACH and composable architecture affect PIM?

MACH stands for microservices based, API-first, cloud-native SaaS and headless. The MACH Alliance launched on 23 June 2020 to promote these principles for enterprise software. In a composable stack, the PIM is one service among several (commerce engine, search, content management, DAM), each connected through APIs.

For PIM selection, composability raises practical criteria:

1. Full API coverage for read and write, including the data model and workflow states.
2. Events or webhooks for change notification.
3. Documented rate limits and bulk endpoints for large catalogues.
4. Authentication standards suited to service-to-service calls.
5. Headless delivery to any front end, with no reliance on a bundled storefront.

## How should DAM and syndication connect?

Assets usually stay in the DAM, and the PIM stores references and selected renditions. That avoids duplicate files and keeps rights metadata in one place, as explained in [DAM and PIM integration](https://pimatlas.com/knowledge/dam-pim-integration/).

Outbound syndication to retailers, marketplaces and data pools such as GDSN often runs through a dedicated syndication layer that maps the PIM model to each recipient's format. See [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/).

## What should an integration design document include?

Before build starts, document each interface with:

- Source, target and owner.
- Pattern (batch, API, event) and frequency or trigger.
- Payload and field mapping, versioned.
- Error handling, retries and alerting.
- Volume estimates and performance targets.
- Cut-over order, which ties into the [PIM data migration](https://pimatlas.com/knowledge/pim-data-migration/) plan.

## Key takeaways

- Decide the system of record per data domain first: ERP for identifiers and logistics, DAM for assets, PIM for enriched content.
- Keep fast-changing price and stock data out of the PIM in most architectures.
- Combine batch, API and event-driven patterns according to freshness needs, and back events with periodic reconciliation.
- MACH principles (microservices, API-first, cloud-native SaaS, headless) translate into concrete API, event and rate-limit criteria.
- Document every interface with owner, pattern, mapping, error handling and volumes before build begins.

Q: Should a PIM integrate directly with ERP or through middleware?
A: Direct point-to-point connectors suit small landscapes with one ERP and one or two channels. Once several systems exchange product data, an integration platform or message broker usually reduces maintenance, because each system connects once and mappings, monitoring and retries sit in one place. The choice depends on volume, change frequency and in-house skills.
Q: What is event-driven PIM integration?
A: In an event-driven integration, the PIM publishes a message when something happens, such as a product being updated or becoming complete for a channel. Subscribed systems react to that message. Akeneo's Event Platform, for example, sends product events in CloudEvents format to Google Pub/Sub, HTTPS endpoints, Kafka or AMQP destinations.
Q: What does MACH mean for PIM?
A: MACH stands for microservices based, API-first, cloud-native SaaS and headless. The MACH Alliance launched in June 2020 to promote these principles. For a PIM, MACH alignment means its data and functions are fully available through APIs, it runs as cloud SaaS and it feeds any front end, so it can be swapped or combined with other composable services.

## Article: What is DAM? Digital asset management for product-led businesses
URL: https://pimatlas.com/knowledge/what-is-dam/
Category: Digital asset management | Updated: 2026-10-17

Summary: A DAM (digital asset management) system is the central library where an organisation stores, describes, approves, transforms and distributes its rich media: product photos, videos, 3D models, manuals and brand files. For product-led businesses, DAM supplies the visual half of every product page, while a PIM supplies the structured data.

A DAM (digital asset management) system is the single source of truth for an organisation's rich media. It stores every approved image, video, 3D model and document once, describes it with metadata, controls who may use it and where, and delivers the correct rendition to each channel. For manufacturers, brands and retailers, the DAM is the visual counterpart to the PIM.

## What does a DAM system do?

A DAM covers the full asset lifecycle, from ingest to archive. The core functions are consistent across vendors, even where the terminology differs.

| Function | What it covers | Typical buyer test |
|---|---|---|
| Ingest | Bulk upload, hot folders, agency portals, API import | Load 5,000 studio images with metadata in one run |
| Metadata and taxonomy | Embedded IPTC/XMP fields, controlled vocabularies, product identifiers | Find every front-facing image for one GTIN |
| Versioning and workflow | Check-in, review, approval, annotation | Route a retouched image to brand approval and back |
| Rights management | Licences, expiry dates, territories, model releases | Block an image after its licence expires |
| Renditions and conversion | On-the-fly resizing, cropping, format conversion | Produce a 1:1.44 JPEG and a 1:1 PNG from one master |
| Distribution | CDN links, portals, connectors to PIM, CMS and commerce | Publish to a marketplace feed without manual export |
| Analytics | Downloads, reuse, channel usage | Show which assets were never used |

The table shows why a shared drive falls short: a DAM adds context, rules and relationships to every stored file.

## How is DAM different from PIM?

A PIM (product information management) system owns structured product data: attributes, descriptions, classifications, variants and relationships. A DAM owns the binary files and their metadata. The two meet at the asset reference, where a product record points to the images, videos and documents that belong to it.

The division of labour is explained in more detail in [PIM vs DAM](https://pimatlas.com/knowledge/pim-vs-dam/). In practice, three questions decide where a piece of content lives:

1. Structured data that varies by product or locale (dimensions, materials, marketing copy) belongs in the PIM.
2. Files with their own lifecycle, versions and rights (photos, video, CAD renders, PDFs) belong in the DAM.
3. Relationships between the two (this image is the primary front view for this GTIN) are stored as links, usually mastered in one system and synchronised to the other.

## Which standards matter for product assets?

Product-led businesses work with external rules that a DAM must support. The GS1 Product Image Standard, released as version 5.0 in July 2026, defines resolutions, backgrounds and a GTIN-based file naming convention for product images. The IPTC Photo Metadata Standard, currently release 2025.1, defines creator, copyright and licensing fields, plus new properties for recording AI prompts and systems. Those fields are embedded in files using XMP, published as ISO 16684-1:2019.

A DAM that reads and writes these standards keeps metadata attached to the file wherever it travels, including after download to an agency, retailer or marketplace. Good [metadata and taxonomy design](https://pimatlas.com/knowledge/dam-metadata-taxonomy/) is therefore the foundation of any DAM project.

## When does a product business need a dedicated DAM?

The basic asset module inside a PIM or commerce platform is often sufficient at first. The following signals indicate that a dedicated DAM is worth evaluating:

- The asset library exceeds roughly 20,000 files or grows by thousands per season.
- More than two teams or external agencies create and upload content.
- Assets carry licences, model releases or territorial restrictions.
- Marketplaces and retailers each require different sizes, ratios and naming.
- Video, 360-degree spins or 3D models are part of the product page.
- Brand and marketing teams reuse the same assets outside product pages.

Several vendors combine both systems in one suite. PIM Atlas profiles DAM-and-PIM suites such as [censhare](https://pimatlas.com/vendors/censhare/), [4ALLPORTAL](https://pimatlas.com/vendors/4allportal/) and [THRON](https://pimatlas.com/vendors/thron/), alongside standalone PIMs that integrate with third-party DAMs.

## What should buyers check first?

Start with the asset types, volumes and channels the business must serve in the next three years. Then test vendors on four practical points: metadata import and export in IPTC/XMP, rendition rules per channel, rights enforcement at the point of download or publication, and the integration pattern with the PIM. A scripted demo using real product assets reveals more than a feature checklist.

## Key takeaways

- A DAM stores, describes, controls and distributes rich media; a PIM manages structured product data.
- Metadata is the core of a DAM: without it, assets are unfindable and unsafe to publish.
- GS1 Product Image Standard 5.0, IPTC Photo Metadata 2025.1 and XMP (ISO 16684-1) are the reference standards for product assets.
- Volume, number of contributors, licensing and channel diversity are the signals that justify a dedicated DAM.
- Evaluate DAM and PIM together, because the asset link between them determines content quality on every channel.

Q: What is the difference between DAM and PIM?
A: A PIM manages structured product data such as attributes, descriptions, classifications and relationships. A DAM manages the media files themselves, with their metadata, versions, renditions and usage rights. Most product pages need both: the PIM holds the facts and the asset references, and the DAM holds the binary files and serves the right rendition to each channel.
Q: Does a small brand need a dedicated DAM?
A: A brand with a few hundred SKUs and one or two channels can often manage with the basic asset module inside its PIM or commerce platform. A dedicated DAM becomes worthwhile when assets run into tens of thousands, several teams or agencies contribute, licences expire, or channels each demand different image sizes, formats and naming rules.
Q: What file types does a DAM manage?
A: A DAM manages any binary file an organisation needs to reuse: product photography, lifestyle images, packaging artwork, video, audio, 3D models in glTF or USDZ, PDF datasheets, safety documents, logos and fonts. Its value comes from the metadata attached to each file, which makes the asset findable, traceable and safe to publish.

## Article: DAM metadata and taxonomy design
URL: https://pimatlas.com/knowledge/dam-metadata-taxonomy/
Category: Digital asset management | Updated: 2026-10-17

Summary: DAM metadata design means deciding which fields describe each asset, which values are allowed and who maintains them. A good model has four layers: technical, descriptive, rights and product context. It embeds standard fields in IPTC and XMP, uses controlled vocabularies, and links every product asset to a GTIN or SKU.

DAM metadata and taxonomy design decides whether assets can be found, trusted and published automatically. The most reliable approach separates metadata into four layers, uses international standards for the fields that already have one, restricts free text with controlled vocabularies, and links every product asset to a product identifier. Design it before migration, because retrofitting metadata on 100,000 files is slow and expensive.

## Which metadata layers does a product DAM need?

A practical model groups fields by purpose and by who maintains them. This keeps ownership clear and prevents one team from editing another team's fields.

| Layer | Example fields | Source | Owner |
|---|---|---|---|
| Technical | Format, pixel dimensions, colour profile, file size, duration | Extracted automatically | DAM system |
| Descriptive | Title, description, keywords, image type, view angle, language | Creator, AI tagging, librarian | Content team |
| Rights | Creator, copyright notice, licence type, expiry date, territories, model release | Contract, photographer | Legal or brand |
| Product context | GTIN, SKU, product family, variant, primary or secondary flag | PIM or ERP | Product data team |
| Lifecycle | Status, approval date, campaign, archive date | Workflow | DAM administrator |

The product context layer is where most product-led DAMs fail. Without a GTIN or SKU on the asset, the [DAM and PIM integration](https://pimatlas.com/knowledge/dam-pim-integration/) depends on file names or manual linking, both of which break at scale.

## Which metadata standards should a DAM support?

Use existing standards for any field they already define. Three are central for product imagery.

- **IPTC Photo Metadata Standard.** The current release is 2025.1, published in November 2025. It defines IPTC Core and IPTC Extension schemas covering creator, credit line, copyright notice, licensor, Web Statement of Rights, people, locations and products shown. Release 2025.1 added four properties for AI prompts and AI systems used.
- **XMP (Extensible Metadata Platform).** XMP is the serialisation format that embeds IPTC and custom fields inside image, PDF and video files. Part 1 is published as ISO 16684-1:2019 and was confirmed in 2024.
- **GS1 Product Image Standard.** Version 5.0 (July 2026) defines a file naming convention that starts with the 14-digit GTIN followed by codes for image type, facing, orientation and packaging state. These codes map directly to descriptive metadata fields.

Embedded standard fields have a direct commercial benefit. Google Images reads Creator, Credit Line, Copyright Notice, Web Statement of Rights and Licensor URL from IPTC metadata and shows them alongside the image.

## How should DAM taxonomies be designed?

A taxonomy is a controlled list of permitted values. In a product DAM, most taxonomies should be reused from elsewhere in the organisation instead of being invented inside the DAM.

1. Reuse the product classification from the PIM for product categories, so a search in the DAM matches the [product taxonomy](https://pimatlas.com/knowledge/product-taxonomy-design/).
2. Define a closed list for image type (packshot, lifestyle, detail, packaging, 360 spin, 3D model, video, document).
3. Use the GS1 facing codes for view angle (front, left, top, back, right, bottom).
4. Keep campaign and season lists time-bound, with an owner who archives old values.
5. Allow free-text keywords only as a supplement to controlled fields.

Flat lists of 10 to 50 values are easier to govern than deep hierarchies. Use hierarchy only where users genuinely browse by level, such as category trees.

## Who owns metadata quality?

Metadata decays without ownership. Assign each layer to a role, set completeness rules per asset type, and block publication when mandatory fields are empty. A packshot without a GTIN or licence expiry should never reach a channel.

AI auto-tagging can populate descriptive keywords quickly, but it should not write rights or product context fields without review. The trade-offs are covered in [AI auto-tagging and visual search in DAM](https://pimatlas.com/knowledge/ai-auto-tagging-dam/). Rights fields deserve the strictest controls, as explained in [managing usage rights in DAM](https://pimatlas.com/knowledge/dam-rights-management/).

Review field usage twice a year. Retire fields with low completion and no search use, and merge synonyms in controlled lists.

## Key takeaways

- Separate metadata into technical, descriptive, rights, product context and lifecycle layers, each with a named owner.
- Use IPTC Photo Metadata 2025.1 fields and embed them with XMP (ISO 16684-1) so information travels with the file.
- Link every product asset to a GTIN or SKU; GS1 Product Image Standard 5.0 naming makes this machine-readable.
- Reuse the PIM classification and closed lists for image type and view angle instead of free text.
- Keep mandatory fields minimal and block publication of assets missing rights or product links.

Q: What is the difference between metadata and taxonomy in a DAM?
A: Metadata is the set of fields attached to each asset, such as creator, GTIN, image type or licence expiry. Taxonomy is the controlled, often hierarchical list of values that some of those fields may take, such as product categories, campaign names or usage types. Metadata gives an asset its description; taxonomy keeps that description consistent across thousands of assets.
Q: Should DAM metadata be embedded in the file or stored in the database?
A: Both. The DAM database holds the authoritative record and supports search and workflow. Core descriptive and rights fields should also be written into the file as IPTC fields in XMP, so creator, copyright and licence information travels with the asset after download. Product-specific fields such as GTIN can be embedded as custom XMP properties if downstream systems read them.
Q: How many metadata fields should a DAM asset have?
A: Keep mandatory fields to the minimum needed for findability, rights and product linkage, typically eight to twelve. Add optional fields only when a named team or channel uses them. Every extra mandatory field slows ingest and lowers completion rates, so measure field usage after six months and retire fields nobody searches or filters on.

## Article: Integrating DAM and PIM: patterns and pitfalls
URL: https://pimatlas.com/knowledge/dam-pim-integration/
Category: Digital asset management | Updated: 2026-10-17

Summary: DAM and PIM integration connects media files to product records so each channel receives the right asset for the right product. Four patterns dominate: a unified suite, reference links with CDN delivery, file copy into the PIM, and event-driven linking by identifier. Reference links with GTIN or SKU matching suit most mid-sized and large catalogues.

The best DAM and PIM integration gives one system clear ownership of each piece of content and lets the other reference it. The PIM owns product data, the DAM owns files and their rights, and a shared identifier (GTIN, SKU or product ID) on every product asset connects the two. Most integration failures trace back to missing identifiers or two systems editing the same link.

## Why integrate DAM and PIM at all?

A product page needs structured data and media at the same moment. When the two live in separate systems without a reliable link, teams compensate with manual uploads, duplicate files and spreadsheets of image names. The consequences are visible on channels: wrong colour variants, expired images still live, and marketplaces rejecting listings for missing or non-compliant pictures.

Integration solves three problems. It removes duplicate storage, it keeps rights and approval status attached to the asset wherever it appears, and it lets the PIM publish complete product records to commerce sites and syndication feeds. The wider context is covered in [PIM integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/).

## Which integration patterns are common?

| Pattern | How it works | Suits | Watch out for |
|---|---|---|---|
| Unified suite | DAM and PIM share one data model and repository | Organisations buying both at once | Weaker depth in one of the two modules |
| Reference link with CDN | PIM stores asset IDs and URLs; DAM serves renditions from a CDN | Most mid-sized and large catalogues | Broken links if assets are deleted or moved |
| File copy | DAM pushes approved files into the PIM asset module | Small catalogues, simple channels | Duplicates, rights status lost after copy |
| Event-driven linking | DAM emits events on approval; middleware matches assets to products by identifier | Complex landscapes with an integration platform | Higher build and monitoring effort |

Several vendors offer a unified suite. PIM Atlas profiles DAM-and-PIM suites such as [censhare](https://pimatlas.com/vendors/censhare/), [4ALLPORTAL](https://pimatlas.com/vendors/4allportal/), [Contentserv](https://pimatlas.com/vendors/contentserv/) and [THRON](https://pimatlas.com/vendors/thron/). Unified suites remove the connector but still require the same governance decisions about ownership and metadata.

## Which system should own the asset link?

Decide this explicitly and document it. A simple rule set works for most organisations:

1. The DAM owns the file, its versions, renditions, rights and approval status.
2. The system where identifiers are assigned at ingest owns the product-to-asset link. For studio photography using GS1 file naming, this is usually the DAM.
3. The PIM owns channel-specific choices: image order, primary image per channel and which assets are excluded from a given marketplace.
4. Neither system deletes an asset that the other still references; expiry triggers a status change and a notification first.

The GS1 Product Image Standard 5.0 naming convention helps here. A file name that begins with the 14-digit GTIN followed by codes for image type, facing and state allows automatic matching without manual tagging.

## What are the most common pitfalls?

- **No identifier on the asset.** Matching by free-text file names fails once suppliers, agencies and studios use different conventions.
- **Rights lost in transit.** Copying a file into the PIM strips the licence expiry, so expired images stay live. Reference links preserve rights checks in the DAM.
- **Rendition sprawl.** Storing every channel size as a separate asset multiplies files. Generate renditions on demand from one master.
- **Two-way editing.** Letting both systems edit metadata such as alt text or captions without a master creates silent overwrites.
- **Variant ambiguity.** Images linked at product family level appear on the wrong colour or size. Link at the level where the visual difference occurs.
- **Ignoring embedded metadata.** IPTC fields written to XMP inside the file keep creator and copyright information intact after download, which database-only metadata cannot do.

Sound [metadata and taxonomy design](https://pimatlas.com/knowledge/dam-metadata-taxonomy/) prevents most of these issues before any integration code is written. During vendor selection, ask each supplier to demonstrate the full path with your own assets: ingest a GS1-named file, match it to a variant, publish it to one channel, then expire its licence and show what happens downstream.

## Key takeaways

- Give each system clear ownership: the DAM owns files and rights, the PIM owns product data and channel choices.
- Put a GTIN, SKU or product ID on every product asset; GS1 file naming makes matching automatic.
- Reference links with CDN delivery suit most catalogues because they preserve rights status and avoid duplicates.
- Unified DAM-and-PIM suites remove the connector but still need the same governance decisions.
- Link images at the variant level where the visual difference occurs, and never let both systems edit the same field.

Q: Should the PIM or the DAM own the link between a product and its images?
A: Either can work, provided only one system masters the link. If photographers and studios assign GTINs at ingest, the DAM usually owns the link and pushes it to the PIM. If product managers choose and order images per channel, the PIM owns the link. Two systems editing the same relationship creates conflicts and missing images.
Q: Do we need a DAM if our PIM already stores images?
A: Not always. A PIM asset module handles packshots for a modest catalogue well. A separate DAM becomes necessary when assets need rights management, approval workflows, many renditions, video and 3D, or reuse by marketing teams outside product pages. At that point the PIM should reference assets held in the DAM instead of storing copies.
Q: How long does a DAM and PIM integration take?
A: With a standard connector and clean identifiers on assets, a first integration typically takes weeks. Custom integrations, legacy assets without GTINs or SKUs, or complex channel rendition rules extend this to several months. The effort is driven more by asset data quality and governance decisions than by the API work itself.

## Article: Product image standards for e-commerce and marketplaces
URL: https://pimatlas.com/knowledge/product-image-standards/
Category: Digital asset management | Updated: 2026-10-17

Summary: Product image standards define the resolution, background, framing, format and file naming that retailers and marketplaces accept. The GS1 Product Image Standard 5.0, ratified in July 2026, is the cross-industry reference. Marketplaces such as Amazon, Zalando and bol add their own rules, so brands should shoot one high-resolution master and generate channel renditions automatically.

Product image standards are the technical rules that decide whether a retailer or marketplace accepts an image. The reference point is the GS1 Product Image Standard, version 5.0, ratified in July 2026. Marketplaces layer their own rules on top, so the practical approach is to shoot one compliant high-resolution master per view and let the DAM generate each channel rendition.

## What does the GS1 Product Image Standard require?

GS1 publishes the standard for images exchanged between brands, retailers and data pools such as GDSN. Release 5.0 renamed the former Product Image Specification and merged in the Product Image Sharing and Delivery Guideline. Release 4.5 (April 2026) had already harmonised sizes and formats and renamed "3D Rendered" images to "3D Model".

The main requirements for primary images are:

- **Resolution:** 900 x 900 to 2400 x 2400 pixels for web primary images, 2401 x 2401 to 4800 x 4800 pixels for high-resolution primary images, and 600 x 600 pixels minimum for optimised mobile hero images.
- **Format:** stored as LZW-compressed TIFF, shared as JPEG or PNG.
- **Background:** white (RGB 255, 255, 255) or transparent, with a clipping path recommended.
- **Angles:** facing codes for front, left, top, back, right and bottom; 360-degree sequences need at least 24 images captured clockwise.
- **File naming:** the 14-digit GTIN, an underscore, then codes for image type, facing, orientation and packaging state.

The standard does not mandate a colour space. It asks for accurate colour without casts, so brands should define their own profile policy, typically sRGB for web delivery.

GTIN-based naming is the most valuable element for product data teams. It lets a DAM or PIM match images to products automatically. The identifier rules themselves are covered in [GS1, GTINs and GDSN explained](https://pimatlas.com/knowledge/gs1-gtin-gdsn/).

## How do marketplace requirements differ?

Each marketplace optimises for its own page layout. The table compares three that matter for European sellers.

| Requirement | Amazon | Zalando | bol |
|---|---|---|---|
| Size | 500 to 10,000 px longest side; over 1,000 px for zoom | Minimum 762 x 1100 px; best practice 1801 x 2600 px | 500 x 500 to 6000 x 6000 px; 1200 x 1200 px for zoom |
| Aspect ratio | Not fixed | Portrait 1:1.44 | Not fixed |
| Format | JPEG preferred; TIFF, PNG, non-animated GIF accepted | JPEG only, sRGB | 30 KB to 10 MB per file |
| Main image background | Pure white RGB 255, 255, 255; product fills 85% of frame | Plain or styled; white packshots placed on light grey | White, neutral, no visible shadows |
| Restrictions | No text, logos or watermarks on main image | Category-specific view counts; CE label image for toys and electronics | No promotional text, logos or watermarks |

Two points stand out. Zalando's portrait ratio means square GS1 masters need a separate crop. bol announced that from 1 July 2025 it no longer enforces its general image guidelines, though it still enforces the rules against self-promotion in images. Marketplace rules change frequently, so assign an owner to monitor them.

## How should a brand build one image set for every channel?

Treat the master as the asset and every channel version as a rendition. A workable sequence:

1. Shoot each view at 2401 to 4800 pixels on the longest side, with GS1 facing codes and enough margin for portrait and square crops.
2. Store a lossless master with a clipping path or transparent background in the DAM.
3. Name or tag each file with GTIN, image type, facing and state, following the GS1 convention.
4. Define rendition rules per channel (size, ratio, background, format, compression).
5. Validate renditions automatically before syndication, rejecting files that break channel rules.

This approach depends on a well-structured DAM. [What is DAM?](https://pimatlas.com/knowledge/what-is-dam/) explains the functions involved, and [DAM metadata and taxonomy design](https://pimatlas.com/knowledge/dam-metadata-taxonomy/) covers the fields that make automated renditions possible.

## How do image standards connect to syndication?

Retailers and data pools increasingly validate images on receipt. A rejected image delays a listing just as surely as a missing attribute. Including image checks in the [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/) workflow, with clear error reporting back to the content team, keeps time to shelf predictable.

## Key takeaways

- The GS1 Product Image Standard 5.0 (July 2026) is the cross-industry reference for resolution, background, angles and GTIN-based file naming.
- Amazon, Zalando and bol each add different size, ratio and background rules; Zalando requires a 1:1.44 portrait ratio.
- Shoot one lossless master per view at 2401 to 4800 pixels and generate channel renditions from it.
- GTIN-based naming lets the DAM and PIM match images to products without manual work.
- Validate images before syndication and assign an owner to track marketplace rule changes.

Q: What is the GS1 Product Image Standard?
A: It is the GS1 global standard for product images shared between brands, retailers and data pools. Version 5.0, ratified in July 2026, merged the former Product Image Specification and the Product Image Sharing and Delivery Guideline. It sets resolution ranges, white or transparent backgrounds, clipping path guidance, facing and angle codes, and a file naming convention that starts with the 14-digit GTIN.
Q: What resolution should product images be?
A: Shoot masters at the high-resolution range in the GS1 standard, 2401 to 4800 pixels on the longest side, and store them losslessly. This covers GS1 primary web images (900 to 2400 pixels), Amazon zoom (over 1,000 pixels on the longest side), bol zoom (1200 by 1200 pixels) and Zalando's recommended 1801 by 2600 pixels.
Q: Do all marketplaces require a white background?
A: No single rule applies. GS1 and Amazon main images require pure white (RGB 255, 255, 255). Zalando accepts plain or styled backgrounds for packshots and automatically places white-background packshots on light grey. bol asks for a white, neutral background without shadows. Keep a clipped master with a transparent or white background so each channel version can be generated.

## Article: Managing usage rights and licences in DAM
URL: https://pimatlas.com/knowledge/dam-rights-management/
Category: Digital asset management | Updated: 2026-10-17

Summary: Rights management in a DAM records who owns each asset, where and until when it may be used, and under which conditions, then enforces those limits automatically. Effective setups capture licence terms at ingest, embed rights fields in IPTC and XMP, block expired assets on every channel, and record AI usage and provenance with IPTC and C2PA.

Rights management in a DAM turns licence contracts into rules the system can enforce. Each asset carries structured data about ownership, permitted channels, territories, time limits and releases, and the DAM blocks use outside those limits. Product businesses that skip this step usually discover the gap through an invoice from a photo agency or a claim from a model's representative.

## Why do product assets need rights management?

Many product images look internally owned but are not. Lifestyle shoots involve photographers, models, locations and stylists, each with contractual limits. Stock imagery carries licence terms. Supplier-provided images may be licensed for resale use only on specific channels. Packaging artwork can include third-party fonts, illustrations or licensed characters.

The risk grows as assets spread. A single lifestyle image may appear on the brand site, three marketplaces, retailer portals, social media and print catalogues. Without central rights data, nobody knows where it is live when the licence ends.

## Which rights data should a DAM record?

Capture rights at ingest, before an asset can be approved. The fields below cover most product and marketing use cases.

| Field | Purpose | Standard |
|---|---|---|
| Creator and credit line | Attribution | IPTC Core |
| Copyright notice and owner | Ownership | IPTC Core and Extension |
| Licensor and Web Statement of Rights | Where to obtain or check the licence | IPTC Extension (PLUS) |
| Licence type | Owned, licensed, royalty-free, rights-managed | Internal vocabulary |
| Permitted channels | Web, marketplaces, social, print, out-of-home | Internal vocabulary |
| Territories | Countries or regions where use is allowed | ISO 3166 country codes |
| Start and expiry dates | Time-bound use | Internal date fields |
| Model and property release | Consent of people and owners shown | IPTC Extension |
| Data mining and AI training | Whether the asset may train AI models | PLUS Data Mining property |

Embedding the IPTC fields in the file as XMP keeps them attached after download. Google Images already reads Copyright Notice, Web Statement of Rights and Licensor URL from this metadata. The wider field model is set out in [DAM metadata and taxonomy design](https://pimatlas.com/knowledge/dam-metadata-taxonomy/).

## How does a DAM enforce rights in practice?

Recording rights achieves little without enforcement. A mature setup applies five controls:

1. **Approval gate.** Assets cannot be approved while mandatory rights fields are empty.
2. **Download and share checks.** Users see restrictions before download, and the DAM blocks channels or territories outside the licence.
3. **Expiry alerts.** Owners receive warnings 30 to 60 days before expiry, with an option to renew.
4. **Automatic withdrawal.** On expiry, CDN-delivered renditions are removed and the PIM receives a status change.
5. **Usage tracking.** The DAM logs where each asset was published so takedowns are complete.

Automatic withdrawal depends on the integration pattern. Reference links let the DAM remove an asset everywhere at once, while copied files need a manual takedown in each system. This is one reason to prefer referencing, as explained in [integrating DAM and PIM](https://pimatlas.com/knowledge/dam-pim-integration/).

## How do AI training and provenance affect rights?

Two developments add new fields to rights management.

The first is reserving rights against text and data mining. Directive (EU) 2019/790 on copyright in the Digital Single Market allows rightholders to reserve their works from commercial text and data mining by machine-readable means. The PLUS Data Mining property, adopted by the IPTC in October 2023, offers a standard embedded field for that reservation, with values that distinguish general data mining, AI or machine learning training and generative AI training.

The second is provenance. The IPTC Photo Metadata Standard 2025.1 added fields for AI prompts and systems used, and the IPTC Digital Source Type vocabulary distinguishes a digital capture from trainedAlgorithmicMedia or compositeWithTrainedAlgorithmicMedia. The C2PA Content Credentials specification goes further, binding a cryptographically signed manifest of creation and edit history to the file. C2PA released version 2.3 in early 2026.

For product content teams, these records support transparency about AI-generated imagery, a topic covered in [the EU AI Act and product content](https://pimatlas.com/knowledge/eu-ai-act-product-content/).

## Key takeaways

- Capture licence terms, territories, channels, dates and releases at ingest, and block approval without them.
- Embed IPTC rights fields in XMP so ownership and licensing information stays with the file.
- Enforce rights through approval gates, expiry alerts, automatic withdrawal and usage tracking.
- Record AI training permissions with the PLUS Data Mining property, consistent with Directive (EU) 2019/790.
- Use IPTC Digital Source Type and C2PA Content Credentials to document whether imagery is captured or AI-generated.

Q: What happens when an image licence expires in a DAM?
A: A well-configured DAM changes the asset status to expired, blocks new downloads, notifies the asset owner and channel teams, and flags every product record and page that still references it. Assets delivered by CDN link can be withdrawn centrally. Assets copied into other systems or downloaded earlier need a manual takedown, which is why reference links are preferred.
Q: Which metadata fields record image rights?
A: The IPTC Photo Metadata Standard provides Creator, Credit Line, Copyright Notice, Copyright Owner, Licensor and Web Statement of Rights, plus Model Release and Property Release status. The PLUS Data Mining property records whether the asset may be used for data mining or AI training. These fields are embedded in the file using XMP so they remain with the asset.
Q: What are Content Credentials and why do they matter for product images?
A: Content Credentials are tamper-evident provenance records defined by the C2PA specification. A signed manifest records who created an asset, which tools edited it and whether generative AI was involved. For product images, they help prove that a packshot is a real capture, document AI retouching, and support transparency expectations for AI-generated marketing content.

## Article: 3D, AR and video product content: managing rich media
URL: https://pimatlas.com/knowledge/3d-ar-video-product-content/
Category: Digital asset management | Updated: 2026-10-17

Summary: Rich media product content covers 360-degree spins, 3D models, augmented reality views and video. Managing it well means storing a source master and delivery formats together: glTF or GLB for web and Android, USDZ for Apple devices, and several video renditions. The DAM governs files and versions; the PIM links each asset to the right product variant.

Rich media product content (360-degree spins, 3D models, augmented reality and video) needs the same governance as still images, plus format management. The working model is one source master per asset, several delivery formats generated from it, and a clear link to the product variant it shows. The DAM manages files, versions and rights; the PIM decides which product and channel each asset belongs to.

## Which rich media types do product pages use?

Each format answers a different buyer question, and each brings its own production and storage demands.

| Type | Buyer question it answers | Delivery format | Production source |
|---|---|---|---|
| 360-degree spin | What does it look like from every side? | Image sequence (JPEG or PNG) | Turntable photography |
| Interactive 3D | How does it fit together, what are the details? | glTF 2.0 / GLB | CAD conversion or 3D scan |
| Augmented reality | How does it look in my room or on my desk? | GLB for Android and web, USDZ for Apple | Same 3D model, optimised |
| Product video | How is it used, installed or maintained? | MP4 renditions, adaptive streaming | Video shoot or animation |
| Configurator render | What does my chosen variant look like? | Rendered images or real-time 3D | 3D model with material library |

The GS1 Product Image Standard 5.0 covers 360-degree imagery: sequences need at least 24 images captured clockwise, with a minimum of 400 x 400 pixels. Release 4.5 renamed the "3D Rendered" image type to "3D Model", which reflects how common rendered packshots have become. See [product image standards](https://pimatlas.com/knowledge/product-image-standards/) for the full specification.

## Which 3D and AR formats matter?

Two delivery formats cover nearly all consumer devices.

- **glTF 2.0 (GL Transmission Format).** A royalty-free format from the Khronos Group for efficient transmission and loading of 3D scenes, published as ISO/IEC 12113:2022 in August 2022. GLB is its single-file binary variant. Google's model-viewer web component displays glTF and GLB in all evergreen browsers and opens AR through Scene Viewer or WebXR on ARCore-supported Android devices.
- **USDZ.** The format Apple uses for AR Quick Look. Safari, Messages, Mail, News and Notes display USDZ models on iPhone, iPad and Apple Vision Pro, and websites can embed Quick Look views.

A common web setup serves a GLB file as the main model and a USDZ file as the iOS alternative from the same product page. Both must come from the same master and carry the same version, or Apple and Android users will see different products.

## How should a DAM manage 3D and video?

Rich media strains DAM systems that were designed for still images. Check these capabilities during selection:

1. **Preview and inspection.** Browser-based 3D and video preview, so reviewers can approve without specialist software.
2. **Asset relationships.** A parent master linked to its GLB, USDZ, texture and poster-image children, versioned together.
3. **Format-specific metadata.** Polygon count, file size, dimensions in real-world units, video duration, captions and language.
4. **Delivery.** CDN delivery with adaptive video streaming and correct MIME types for GLB and USDZ.
5. **Size limits.** Upload and storage limits that handle multi-gigabyte CAD and video masters.
6. **Rights.** Music licences, voice-over rights and model releases for video, recorded like any other licence.

The foundations are the same as for images, covered in [What is DAM?](https://pimatlas.com/knowledge/what-is-dam/).

## How do rich media assets link to products?

Link rich media at the level where the visual difference occurs. A 3D model of a sofa in grey fabric is wrong for the blue variant, while an installation video often applies to a whole product family. The PIM should record the role of each asset (spin, 3D, AR, demo video, installation video) so channels can request the right type.

For manufacturers, CAD-derived 3D is often the most efficient source, provided internal geometry is removed before publication. The handover from engineering is described in [PIM for manufacturers](https://pimatlas.com/knowledge/pim-for-manufacturers/).

## Key takeaways

- Keep one source master per rich media asset and generate delivery formats from it.
- Use glTF 2.0 (ISO/IEC 12113:2022) as GLB for web and Android, and USDZ for Apple AR Quick Look.
- GS1 Product Image Standard 5.0 sets at least 24 frames and 400 x 400 pixels for 360-degree spins.
- Test DAMs on 3D and video preview, parent and child asset relationships, large file handling and rights for video.
- Link rich media to the product variant it actually shows, and record each asset's role in the PIM.

Q: Which 3D file format should we use for e-commerce?
A: Use glTF 2.0, usually as a single binary GLB file, for web viewers and Android AR. It is a royalty-free Khronos format published as ISO/IEC 12113:2022. Add a USDZ version for AR Quick Look on iPhone, iPad and Apple Vision Pro. Keep the original CAD or authoring file as the master in the DAM and generate both delivery formats from it.
Q: Can product 3D models be created from CAD data?
A: Yes, and for manufacturers this is often the most efficient route. Engineering CAD files are converted, simplified to a manageable polygon count and given realistic materials before export to glTF and USDZ. The conversion needs care: CAD models may contain confidential internal geometry, and the visual model should show only what the customer sees.
Q: Where should product videos be stored: DAM, PIM or video platform?
A: Store the master file, metadata, rights and approval status in the DAM. Use a streaming platform or the DAM's video delivery service for adaptive playback renditions and captions. The PIM should hold only a reference to the video and its role, such as installation guide or product demo, linked to the relevant products and variants.

## Article: The product content supply chain: from creation to channel
URL: https://pimatlas.com/knowledge/product-content-supply-chain/
Category: Digital asset management | Updated: 2026-10-17

Summary: The product content supply chain is the end-to-end process that turns a new product into publishable data and media on every channel. It runs through six stages: plan, source, create, enrich, approve and distribute. PIM and DAM provide the backbone, and shared identifiers, clear handovers and stage-level metrics keep content moving.

The product content supply chain is the end-to-end flow that turns a new product into complete, approved data and media on every channel where it is sold. Organisations that manage it as a supply chain, with standard inputs, defined handovers and measured lead times, launch products faster and with fewer rejections. PIM and DAM are its backbone, linked by a shared product identifier.

## What are the stages of the product content supply chain?

Most organisations can map their process onto six stages. Each has a distinct owner, system and output.

| Stage | Main activities | Primary system | Output |
|---|---|---|---|
| Plan | Assortment decisions, content briefs, channel requirements | PLM, ERP, PIM | Product record with identifiers |
| Source | Supplier data, CAD files, packaging artwork, regulatory documents | Supplier portal, PIM, DAM | Raw data and files linked to GTIN or SKU |
| Create | Photography, 3D, video, copywriting | DAM, studio tools | Asset masters and draft copy |
| Enrich | Attributes, classifications, translations, SEO copy, asset linking | PIM | Channel-ready product record |
| Approve | Brand, legal, regulatory and quality checks | PIM and DAM workflows | Approved data and assets |
| Distribute | Commerce sites, marketplaces, retailers, print, data pools | PIM, syndication, CDN | Live product content |

The stages overlap in practice. Photography often starts before supplier data is complete, and enrichment continues after launch. The table still helps assign ownership and locate delays.

## Where do PIM and DAM fit?

The PIM carries the product record through the chain: it holds identifiers, attributes, copy and channel rules, and it decides when a product is complete for a channel. The DAM carries the assets: masters, renditions, rights and approval status. The link between the two, ideally a GTIN or SKU embedded in each asset, is the point where most supply chains break.

Some vendors combine both in one platform. PIM Atlas profiles DAM-and-PIM suites such as [censhare](https://pimatlas.com/vendors/censhare/) and [Contentserv](https://pimatlas.com/vendors/contentserv/). Others integrate separate systems; the trade-offs are covered in [integrating DAM and PIM](https://pimatlas.com/knowledge/dam-pim-integration/).

## What causes bottlenecks?

Bottlenecks tend to appear at handovers. The most frequent are:

- **Late or incomplete supplier data.** Enrichment waits for basic attributes and dimensions.
- **Photography backlog.** Samples arrive late, or shot lists lack GTINs, so images cannot be matched to products.
- **Translation queues.** Copy is translated after approval in one language, adding weeks per market.
- **Serial approvals.** Brand, legal and compliance review one after another instead of in parallel.
- **Channel rejections.** Images or attributes fail retailer validation and loop back to earlier stages.

Each bottleneck has a structural fix. Supplier onboarding templates address the first; GS1 file naming with the 14-digit GTIN at the shoot addresses the second; parallel workflows and channel validation before syndication address the rest. [Product data governance](https://pimatlas.com/knowledge/product-data-governance/) sets the roles and rules that make these fixes stick.

## How should the supply chain be measured?

Measure the chain at stage level, using a small set of indicators:

1. **Lead time to first channel:** days from product creation in the PIM to the first live listing.
2. **Channel completeness:** the share of products meeting each channel's mandatory data and asset rules.
3. **First-time acceptance rate:** the share of submissions accepted by retailers and marketplaces without correction.
4. **Rework loops:** average number of times a product returns to an earlier stage.
5. **Asset reuse rate:** the share of DAM assets used on more than one channel.

Completeness and acceptance metrics overlap with [product data quality](https://pimatlas.com/knowledge/product-data-quality/) scoring, so define them once and share them across teams. Review the figures monthly with every stage owner, and set targets per product category, because a fashion range and a spare-parts catalogue have very different lead times.

## How does AI change the supply chain?

Generative and analytical AI speeds up the create and enrich stages: drafting descriptions, translating, extracting attributes and tagging assets. It also adds provenance duties. The IPTC Digital Source Type vocabulary distinguishes a digital capture from trainedAlgorithmicMedia, and C2PA Content Credentials, at version 2.3 since early 2026, record signed edit histories. Building these records into the approve stage keeps AI-assisted content traceable.

## Key takeaways

- The product content supply chain runs through six stages: plan, source, create, enrich, approve and distribute.
- PIM carries the product record, DAM carries the assets, and a shared GTIN or SKU links them.
- Most delays occur at handovers: supplier data, photography matching, translation and serial approvals.
- Measure lead time, channel completeness, first-time acceptance, rework and asset reuse per stage.
- Record AI involvement with IPTC Digital Source Type or C2PA Content Credentials during approval.

Q: What is a content supply chain in product information management?
A: It is the sequence of steps, people and systems that produce and deliver product content, from supplier data and photography through enrichment, translation and approval to publication on websites, marketplaces and print. The term borrows from physical supply chains because the same principles apply: standard inputs, clear handovers, measured lead times and quality checks at each stage.
Q: How do you measure a product content supply chain?
A: Track lead time from product creation to first live channel, the share of products complete for each channel, first-time acceptance rate at retailers and marketplaces, rework loops per product, and asset reuse rate. Measure these per stage, so bottlenecks such as photography backlogs or translation queues become visible and can be fixed with capacity or process changes.
Q: Where does generative AI fit in the product content supply chain?
A: Generative AI mostly accelerates the create and enrich stages: drafting descriptions, translating, extracting attributes and tagging images. It increases throughput but adds review and provenance requirements. Teams should keep human approval for regulated content, label AI-generated imagery with IPTC Digital Source Type or C2PA Content Credentials, and track AI output quality as a separate metric.

## Article: AI in PIM: what delivers value in 2026
URL: https://pimatlas.com/knowledge/ai-in-pim/
Category: AI and search | Updated: 2026-10-17

Summary: AI in PIM delivers measurable value in 2026 in four areas: extracting attributes from supplier documents, classifying products, translating content and drafting copy under review. Agents and MCP servers now let external AI tools read and update PIM data. Results depend on a clean data model, clear validation rules and human approval steps.

AI in a PIM (product information management) system now delivers dependable value in four jobs: extracting attributes from unstructured supplier data, classifying products into a taxonomy, translating content and drafting copy for human review. The newer layer, AI agents and MCP servers, lets external assistants work on PIM data directly. Every one of these depends on the quality of the underlying data model.

## Which AI capabilities have PIM vendors shipped?

By October 2026 almost every enterprise and mid-market PIM vendor markets AI features, and most now describe them as agents. Naming varies, so buyers should map vendor terms to the underlying jobs.

| Job | What the AI does | Examples as vendors name them |
|---|---|---|
| Supplier onboarding | Reads PDFs, spreadsheets and catalogues, maps values to attributes | Akeneo AI Extraction (Supplier Data Manager), inriver Inspire onboarding |
| Classification | Assigns category and taxonomy codes | Syndigo AI GoPilots classification and taxonomy mapping |
| Enrichment and copy | Fills attributes, drafts descriptions and bullets | Akeneo Text and Image Enrichment, Pimcore enrichment and content agents, Plytix AI content studio |
| Translation | Localises content with glossaries | inriver Inspire translation, Pimcore translation agents |
| Quality monitoring | Flags gaps, anomalies and inconsistencies | Akeneo Agentic Product Intelligence, Syndigo AI GoPilots anomaly detection |
| Agent access | Exposes PIM data to external AI tools via MCP | Stibo Systems MCP Server, Bluestone PIM MCP (Agent Gateway), inriver Inspire MCP server |

Salsify groups its features under Salsify Intelligence Suite and the Angie assistant, Stibo Systems markets AgentWorkx as its engine for agentic MDM, and Bluestone PIM lists AI Enrich, AI Linguist, AI Analyst, AI DAM and AI Workflow. The labels change quickly, so ask for a live demonstration on your own data.

## Where does AI deliver measurable value?

Value is easiest to prove where a task is repetitive, rule-bound and verifiable. Attribute extraction from a supplier datasheet can be checked against the source. A classification can be checked against the taxonomy. A translation can be checked against a glossary. These jobs produce throughput metrics (records per hour, first-pass acceptance rate) that support a [business case for PIM](https://pimatlas.com/knowledge/pim-business-case/).

Generative copy delivers value when it works from structured attributes and brand rules. Google's guidance warns that mass-produced pages without added value can breach its spam policy on scaled content abuse, and that all AI output should be fact-checked before publication. Our guide to [generative AI for product descriptions](https://pimatlas.com/knowledge/generative-ai-product-descriptions/) covers review workflows in detail.

## What do agents and MCP change?

The Model Context Protocol (MCP) is an open-source standard for connecting AI applications to external systems, now governed under the Agentic AI Foundation at the Linux Foundation. A PIM with an MCP server lets assistants such as ChatGPT or Claude query products, read completeness scores or propose edits under the permissions the PIM grants.

Two consequences follow. Product teams can run bulk analysis in natural language without exporting data. External shopping agents and retailer systems can also consume governed product data more directly, which links PIM to [agentic commerce](https://pimatlas.com/knowledge/agentic-commerce-product-data/). Both raise the stakes on access control and audit logging.

## How should buyers evaluate AI claims?

Treat AI as a set of requirements in the RFP and score each one on evidence. Useful criteria:

1. **Model choice and hosting.** Which large language models are used, where data is processed and whether customer data trains shared models.
2. **Grounding.** Whether generation draws only on PIM attributes, assets and approved reference text.
3. **Human review.** Whether AI changes land as drafts in a workflow with approval, rejection and versioning.
4. **Validation.** Whether outputs pass the same completeness and format rules as manual edits.
5. **Audit trail.** Whether each value records its origin (human, import, AI) and the prompt or agent used.
6. **Cost model.** Whether AI usage is bundled, metered by credits or billed per call.

Add these to your [PIM RFP requirements](https://pimatlas.com/knowledge/pim-rfp-requirements/) and test them in scripted demos with your own records.

## What has to be in place first?

AI amplifies the state of the data. A PIM with inconsistent attribute definitions, missing units and no ownership will produce inconsistent AI output faster. Before scaling AI, organisations should have a documented data model, defined [product data quality](https://pimatlas.com/knowledge/product-data-quality/) metrics and named data owners. Labelling of AI-generated values also matters for channels: Google Merchant Center asks merchants to submit AI-generated titles and descriptions in separate attributes.

## Key takeaways

- AI in PIM delivers proven value in extraction, classification, translation and drafted copy.
- Agents and MCP servers let external AI tools read and act on PIM data, which raises governance requirements.
- Map vendor feature names to underlying jobs and test them on your own data.
- Score AI on model hosting, grounding, review workflow, validation, audit trail and cost.
- A clean data model and clear ownership come before any AI rollout.

Q: Which AI features in a PIM system deliver the fastest return?
A: Attribute extraction from supplier files, automated classification and machine translation usually pay back first, because they replace repetitive manual work with measurable throughput. Generative copy follows once tone, terminology and approval workflows are defined. Agent features depend on a stable data model and governance, so they tend to come later in the roadmap.
Q: What is an MCP server in a PIM?
A: An MCP server exposes PIM data and actions through the Model Context Protocol, an open standard for connecting AI applications to external systems. It lets assistants such as ChatGPT or Claude query products, read attributes or trigger updates under the permissions the PIM grants, without a custom integration for each AI tool.
Q: Do I need a new PIM to use AI on product data?
A: Not necessarily. Many current PIM platforms ship AI features or connect to external models through APIs and MCP. Replacement makes sense when the existing data model, workflow engine or API cannot support validation, review steps and audit trails for AI-generated changes. Assess the data foundation before the AI feature list.

## Article: Generative AI for product descriptions: quality and governance
URL: https://pimatlas.com/knowledge/generative-ai-product-descriptions/
Category: AI and search | Updated: 2026-10-17

Summary: Generative AI writes usable product descriptions when it is grounded in structured PIM attributes, constrained by brand and terminology rules, and routed through human review. The main risks are invented facts, repetitive copy at scale and missing labels. Treat every generated description as a draft with a recorded source, prompt and approver.

Generative AI produces usable product descriptions when three conditions hold: the model writes only from structured product attributes, brand and terminology rules constrain the output, and a person approves each description before it is published. Without these controls, generative AI scales errors and repetitive copy as fast as it scales output.

## Where does generative AI fit in the description workflow?

A product description in a PIM (product information management) system is a derived field: it should restate and explain facts already held as attributes. That makes it a good fit for a large language model (LLM), which can turn a list of specifications into readable prose in many languages and channel formats.

Most PIM vendors now ship this capability. Akeneo markets Text and Image Enrichment, inriver positions its Inspire engine to create and refine product copy with brand guidelines as prompt-level guardrails, and Plytix offers an AI content studio. See [AI in PIM](https://pimatlas.com/knowledge/ai-in-pim/) for a wider map of features.

## What are the main quality risks?

| Risk | Example | Control |
|---|---|---|
| Invented facts | Claims a jacket is waterproof when the attribute says water-resistant | Ground on attributes only, automated fact check |
| Missing data hidden | Model writes around an empty dimension field | Block generation until required attributes are complete |
| Repetition at scale | 5,000 descriptions with identical openings | Variation rules, sample review per batch |
| Tone drift | Copy breaks brand voice or uses banned claims | Style guide in prompt, banned-term filter |
| Regulated claims | Health, safety or environmental claims without evidence | Claims whitelist, legal review for flagged categories |
| Translation errors | Technical terms mistranslated | Glossaries, native-speaker review for key markets |

Google's guidance states that generative models can produce inaccurate output and recommends fact-checking before publishing. It also warns that mass-produced pages without added value can breach its spam policy on scaled content abuse.

## How should a governed workflow run?

A governed workflow treats every generated description as a draft. The steps below fit most PIM workflow engines:

1. **Gate on completeness.** Generate only when the record passes [product data quality](https://pimatlas.com/knowledge/product-data-quality/) thresholds for the category.
2. **Assemble the prompt from the record.** Include attributes, units, target channel, language, length limit and style rules. Exclude anything the model should not use.
3. **Generate and validate.** Check length, banned terms and that every number in the output matches an attribute value.
4. **Review.** A named editor approves, rejects or edits. High-risk categories need a second reviewer.
5. **Record provenance.** Store the model, prompt version, reviewer and date with the value.
6. **Publish with labels where required.** Map the description to the correct channel attribute.

Assign ownership for prompts, glossaries and review in your [product data governance](https://pimatlas.com/knowledge/product-data-governance/) model, and version prompts like any other configuration.

## What labelling and legal rules apply?

Google Merchant Center asks merchants to submit AI-generated titles and descriptions through the structured_title and structured_description attributes, with the digital_source_type value trained_algorithmic_media. The content limits match the standard attributes: 150 characters for titles and 5,000 for descriptions.

In the EU, Article 50 of the AI Act (Regulation (EU) 2024/1689) applies from 2 August 2026. Deployers must disclose AI-generated text published to inform the public on matters of public interest, unless it has undergone human review or editorial control. Routine product copy is generally outside that duty, though providers of generative systems must mark outputs in a machine-readable format. Our guide to the [EU AI Act for product content teams](https://pimatlas.com/knowledge/eu-ai-act-product-content/) covers the detail.

## How do you measure description quality?

Measure output against criteria a reviewer can apply consistently:

- **Factual accuracy:** share of descriptions with zero unsupported claims in sample review.
- **First-pass acceptance:** share approved without edits, tracked by category and language.
- **Coverage:** share of active SKUs with a channel-ready description per market.
- **Distinctiveness:** share of descriptions with unique opening sentences within a category.
- **Channel compliance:** rejection rate from marketplaces and feeds.

These metrics show where prompts, source data or review capacity need attention, and support decisions to widen or narrow automation per category.

## Key takeaways

- Ground generation in structured PIM attributes and block it when required data is missing.
- Treat every AI description as a draft with a recorded model, prompt and approver.
- Automated checks for numbers, banned terms and length catch most errors before review.
- Use Google Merchant Center's structured_title and structured_description for AI-generated text.
- Measure accuracy, acceptance rate and distinctiveness per category.

Q: Can AI-generated product descriptions hurt search rankings?
A: Google does not penalise AI-generated content as such. Its guidance warns that generating many pages without adding value can breach the spam policy on scaled content abuse. Descriptions grounded in accurate, product-specific attributes and reviewed by people meet the same standards as manually written copy.
Q: Do I have to label AI-written product descriptions?
A: Google Merchant Center asks merchants to submit generative AI titles and descriptions through structured_title and structured_description with the trained_algorithmic_media value. Under EU AI Act Article 50, the text disclosure duty targets public-interest publications, so ordinary product copy is generally outside it, but check consumer law and marketplace rules.
Q: How do I stop AI inventing product features?
A: Ground the prompt in PIM attributes only, instruct the model to use no facts outside them, and run automated checks that every number, material and certification in the output exists in the source record. Route outputs through an approval step, reject drafts that fail the check and fix the prompt or source data that caused the error.

## Article: AI enrichment: attribute extraction, classification and translation
URL: https://pimatlas.com/knowledge/ai-product-data-enrichment/
Category: AI and search | Updated: 2026-10-17

Summary: AI enrichment fills product records automatically in three ways: extracting attribute values from supplier documents and images, assigning products to taxonomy classes such as ETIM or GS1 GPC, and translating content with controlled glossaries. It works best with confidence thresholds, human review of low-confidence values and accuracy measured against a labelled sample.

AI product data enrichment fills and standardises product records with three techniques: extracting attribute values from unstructured sources, classifying products into a taxonomy, and translating content. Each technique works reliably when it outputs a confidence score, low-confidence values go to a human reviewer and accuracy is measured against a labelled sample of real records.

## What is AI product data enrichment?

Product data enrichment is the process of adding missing attributes, classifications, descriptions and translations to a product record so it is complete for each channel. AI enrichment uses machine learning models, mostly large language models (LLMs) and vision models, to do part of that work automatically inside or alongside a PIM (product information management) system.

It differs from generative copywriting, covered in [generative AI for product descriptions](https://pimatlas.com/knowledge/generative-ai-product-descriptions/), because the output is structured data: a value in a defined attribute, a class code or a translated field.

## How does attribute extraction work?

Attribute extraction reads unstructured sources (supplier PDFs, datasheets, spreadsheets, web pages, packaging images) and maps values to attributes in the target data model. A model identifies that "IP65" belongs in an ingress protection attribute, normalises "1,2 kg" to 1.2 kilograms and flags conflicting values across sources.

Vendors ship this mainly for supplier onboarding. Akeneo includes AI Extraction in its Supplier Data Manager to extract, map and normalise supplier information. inriver describes multimodal onboarding in its Inspire engine that ingests multiple file formats and validates entities. For distributors handling thousands of suppliers, this is often the highest-volume AI use case, as described in [PIM for wholesale and distribution](https://pimatlas.com/knowledge/pim-for-wholesale-distribution/).

Vision models extend extraction to images. Google Cloud Vision label detection, for example, returns labels with a confidence score from 0 to 1, which can populate colour, pattern or product-type attributes. Image-derived values need the same review rules as text-derived ones, because packaging shots and lifestyle images often show accessories that are not part of the product.

## How does automated classification work?

Classification assigns each product to a class in a taxonomy: an internal category tree, GS1 Global Product Classification (GPC), ETIM or ECLASS. The model reads the product name, description and any attributes, then proposes one or more classes with a confidence score.

Classification has a multiplier effect. In [ETIM and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/) the class defines which features a product must carry, so a wrong class produces a wrong attribute set. Syndigo's AI GoPilots, for example, include auto-classification against predefined taxonomies and alignment with standards such as GS1 and ETIM. A clear [product taxonomy design](https://pimatlas.com/knowledge/product-taxonomy-design/) with written class definitions improves model accuracy.

## How should translation be automated?

Machine translation with LLMs now handles most descriptive product content. Quality depends on controls around the model:

- **Glossaries** for technical terms, brand names and protected product names.
- **Unit and format rules** per locale (decimal separators, measurement systems, date formats).
- **Do-not-translate lists** for model numbers, standards and trademarks.
- **Review tiers** that send safety, legal and regulated text to fluent reviewers.

inriver and Pimcore both describe translation agents, with inriver citing custom glossaries for consistent terminology.

## How do you measure enrichment accuracy?

Measure accuracy before scaling. A practical method is to label a sample of 200 to 500 records by hand, run the AI on the same records and compare results attribute by attribute.

| Metric | Definition | Use |
|---|---|---|
| Precision | Share of AI values that are correct | Sets the confidence threshold for auto-acceptance |
| Coverage | Share of target attributes the AI fills | Shows where source data is too thin |
| Review rate | Share of values routed to humans | Sizes the review team |
| Classification accuracy | Share of products in the correct class | Tracked per taxonomy level |
| Translation error rate | Errors per 1,000 words by reviewer | Tracked per language |

Confidence scores turn these metrics into rules. Values above a threshold set per attribute can be accepted automatically. Values below it go to a review queue. Thresholds should be stricter for attributes with legal or safety impact, such as dimensions, voltages or allergens.

Record the origin of every enriched value (AI, supplier, human) so that [product data quality](https://pimatlas.com/knowledge/product-data-quality/) reports can separate AI performance from source data problems.

## Key takeaways

- AI enrichment covers attribute extraction, classification and translation, each producing structured data.
- Attribute extraction is the highest-volume use case for distributors and retailers onboarding supplier data.
- Classification errors cascade into attribute sets, so review low-confidence proposals.
- Glossaries, unit rules and review tiers determine machine translation quality.
- Measure precision and coverage on a labelled sample before setting auto-acceptance thresholds.

Q: How accurate is AI attribute extraction from supplier PDFs?
A: Accuracy depends on document quality, attribute type and how well the target data model is defined, so measure it on your own documents. Run a pilot on a labelled sample of a few hundred records, measure precision per attribute, and set confidence thresholds so that values below the threshold go to human review instead of straight into the PIM.
Q: Can AI classify products into ETIM or ECLASS automatically?
A: Yes, AI models can propose a class from a product name, description and attributes, and several PIM vendors ship automated classification. Because ETIM and ECLASS classes determine which features a product must carry, a wrong class cascades into wrong attributes. Review low-confidence proposals and new suppliers before accepting them.
Q: Is machine translation good enough for product data?
A: Machine translation handles most descriptive product text well when supported by glossaries for technical terms, units and brand names. Safety information, legal claims and regulated content need human review by a fluent speaker. Track error rates per language and category, and prioritise review for markets with the highest revenue or regulatory exposure.

## Article: Generative engine optimisation (GEO) for product content
URL: https://pimatlas.com/knowledge/geo-product-content/
Category: AI and search | Updated: 2026-10-17

Summary: Generative engine optimisation (GEO) for product content means making product facts easy for AI assistants to find, trust and quote. In practice that requires complete structured attributes, consistent identifiers such as GTINs, schema.org Product markup, current merchant feeds and answer-ready content such as specifications, compatibility and FAQs, all published from one governed PIM record.

Generative engine optimisation (GEO) for product content is the practice of making product facts easy for AI assistants and AI search features to retrieve, trust and quote. For product data, GEO rests on four foundations: complete structured attributes, consistent identifiers, accurate structured data and feeds, and content that answers specific shopper questions.

## What is GEO and how does it apply to products?

The term comes from a 2023 research paper, "GEO: Generative Engine Optimization" by Aggarwal and colleagues (arXiv 2311.09735), which reported that targeted content changes can raise visibility in generative engine responses by up to 40%, with effects varying by domain. Generative engines include ChatGPT, Gemini, Microsoft Copilot, Perplexity and Google's AI Overviews and AI Mode.

For products, an assistant typically answers a question such as "quiet dishwasher under 45 dB for a small kitchen" by combining web pages, merchant feeds and its own index. Products with clear, comparable facts are easier to select and describe correctly. GEO for products is therefore mostly a [product data quality](https://pimatlas.com/knowledge/product-data-quality/) discipline.

## Which signals do AI assistants use?

| Signal | Where it comes from | What to manage in PIM |
|---|---|---|
| Identifiers | GTIN, brand, MPN in pages and feeds | One verified GTIN per trade item, consistent brand name |
| Core facts | Specifications, dimensions, materials | Typed attributes with units, no facts only in prose |
| Commercial data | Price, availability, shipping, returns | Synchronised feeds, frequent updates |
| Variants | Size, colour, configuration | Item group and variant options modelled explicitly |
| Context | Use cases, compatibility, FAQs | Dedicated attributes for compatibility and Q&A |
| Evidence | Reviews, certifications, documents | Linked manuals, certificates, test results |

Consistent identifiers matter because assistants merge information from many sources. A product that carries different names or GTINs across retailers is harder to reconcile, as explained in our guide to [GS1 and GTINs](https://pimatlas.com/knowledge/gs1-gtin-gdsn/).

## What does Google say about AI search visibility?

Google's Search Central documentation states that there are no additional requirements or special optimisations needed to appear in AI Overviews or AI Mode, and that no new machine-readable or AI text files are required. Pages need to be indexed and eligible for a snippet, and structured data should match visible content.

For shopping, Google recommends pairing schema.org Product markup on pages with a Merchant Center feed. The schema.org Product type supports gtin, mpn, sku, brand, offers, additionalProperty, hasMerchantReturnPolicy and isVariantOf with ProductGroup for variants.

At Google Marketing Live on 20 May 2026, Google announced conversational attributes in Merchant Center, now documented as question_and_answer, document_link, related_product, item_group_title, variant_option and popularity_rank. They are optional, can be supplied through a supplemental data source and are intended to help products appear on AI-driven surfaces such as AI Mode.

## How should product content be written for AI answers?

Content that AI assistants can quote has a clear structure and measurable claims:

1. **Lead with the defining facts.** Product type, key specification and primary use in the first sentence.
2. **State numbers with units.** "Noise level 42 dB(A)" is quotable. "Whisper-quiet" is not.
3. **Answer real questions.** Use customer service logs, reviews and return reasons to write Q&A pairs.
4. **Describe compatibility explicitly.** List compatible models, accessories and replacement parts as structured relationships.
5. **Keep one version of each fact.** Price, dimensions and claims must match across site, feeds and retailers.
6. **Link supporting documents.** Manuals, safety sheets and certificates add verifiable evidence.

## How does PIM support GEO?

A PIM (product information management) system is the natural home for GEO work because it already holds the governed record that feeds websites, marketplaces and retailers. Practical steps include adding attributes for Q&A, compatibility and use cases, mapping them to schema.org properties and Merchant Center attributes, and monitoring completeness per channel. [Product data syndication](https://pimatlas.com/knowledge/product-data-syndication/) then distributes the same facts to retailers and marketplaces.

The same data serves AI shopping agents that act on behalf of buyers. Our guide to [agentic commerce](https://pimatlas.com/knowledge/agentic-commerce-product-data/) covers the protocols and feeds those agents use.

## Key takeaways

- GEO for products depends on complete, consistent, structured product facts.
- Google requires no special markup for AI Overviews or AI Mode, so standard structured data and feeds remain the foundation.
- Merchant Center conversational attributes add Q&A, documents, related products and variant details for AI surfaces.
- Write content with measurable facts, explicit compatibility and answers to real questions.
- Manage GEO attributes in PIM and syndicate one version of each fact to every channel.

Q: What is the difference between SEO and GEO for products?
A: SEO aims to rank product pages in search results. GEO aims to have products selected, described and cited inside answers from AI assistants and AI search features. Both rely on crawlable, accurate pages, but GEO places more weight on structured facts, consistent identifiers, merchant feeds and content that answers specific shopper questions.
Q: Does Google require special markup for AI Overviews or AI Mode?
A: No. Google states there are no additional requirements or special optimisations needed to appear in AI Overviews or AI Mode, and no new machine-readable or AI text files are required. Standard structured data should match visible page content. For shopping surfaces, Merchant Center feeds and the optional conversational attributes add product detail.
Q: Which product attributes matter most for AI assistants?
A: Identifiers (GTIN, brand, MPN), price and availability, core specifications with units, variant options, compatibility and use cases, and answers to common questions. Assistants compare products on these facts, so missing or inconsistent values reduce the chance of being recommended. Marketing adjectives without measurable facts add little.

## Article: Agentic commerce: preparing product data for AI shopping agents
URL: https://pimatlas.com/knowledge/agentic-commerce-product-data/
Category: AI and search | Updated: 2026-10-17

Summary: Agentic commerce is buying in which an AI agent discovers, compares and purchases products for a person. Agents rely on structured product feeds and open protocols such as OpenAI's Agentic Commerce Protocol, Google's Universal Commerce Protocol and Agent Payments Protocol. Preparation starts in the PIM: identifiers, complete attributes, variants, current price and availability.

Agentic commerce is shopping in which an AI agent searches, compares and, with permission, buys products on a person's behalf. Agents cannot read a product page the way a person does, so they depend on structured product feeds and open protocols. Product data teams prepare by making identifiers, attributes, variants, price and availability complete, consistent and current.

## What is agentic commerce?

An AI shopping agent is software built on a large language model that takes a goal ("find white running shoes under 120 euros in size 44") and carries out the steps: discovery, comparison, cart and checkout. Between September 2025 and May 2026, the major AI and commerce platforms published protocols that define how agents, merchants and payment providers exchange data.

The product data requirement is constant across them: an agent can only recommend and buy what it can identify and compare.

## Which protocols and programmes matter?

| Protocol or programme | Announced | Backers | Role |
|---|---|---|---|
| Agentic Commerce Protocol (ACP) | 29 September 2025 | OpenAI, Stripe | Product feeds, checkout and delegated payment for ChatGPT |
| Agent Payments Protocol (AP2) | 16 September 2025 | Google, more than 60 organisations | Signed Intent and Cart Mandates proving what the user authorised |
| Universal Commerce Protocol (UCP) | 11 January 2026 | Google with Shopify, Etsy, Wayfair, Target, Walmart | Discovery to order management, compatible with AP2 |
| Shopify Catalog and Agentic Storefronts | 11 January 2026 | Shopify | Syndicates merchant products to ChatGPT, Microsoft Copilot, Google AI Mode and Gemini |
| Model Context Protocol (MCP) | Open standard | Agentic AI Foundation (Linux Foundation) | Connects AI applications to external systems and data |

The landscape is moving. OpenAI launched Instant Checkout for US Etsy sellers in September 2025. In March 2026 OpenAI said Instant Checkout was moving to retailer apps inside ChatGPT, while ACP development continued. At Google Marketing Live on 20 May 2026, Google announced Universal Cart across Search and Gemini and UCP checkout in ads. Plan for several protocols in parallel.

## What do AI shopping agents need from product data?

Feed specifications show what agents consume. OpenAI's product feed specification requires item_id, title, description, url, brand, seller_name, image_url, availability and price, with gtin and mpn optional in its own format and required in its Google-compatible format unless no identifier exists. Flags such as is_eligible_search and is_eligible_checkout control where products appear. Google's UCP implementation requires an active Merchant Center account with eligible products.

A practical readiness checklist:

1. **Identifiers.** A valid GTIN per trade item, plus brand and MPN. See [GS1, GTINs and GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/).
2. **Comparable attributes.** Specifications with units, materials, dimensions and certifications as typed fields.
3. **Variant structure.** Item groups with explicit options for size, colour and configuration.
4. **Commercial freshness.** Price, availability and shipping updated whenever they change.
5. **Policies.** Returns, warranty and delivery terms in machine-readable form.
6. **Context.** Compatibility, use cases and Q&A, which also serve [generative engine optimisation](https://pimatlas.com/knowledge/geo-product-content/).
7. **Images.** Direct image URLs per variant that meet channel requirements.

## How does PIM fit into agentic commerce?

A PIM (product information management) system holds the governed product record, while commerce, ERP and order systems hold live price, stock and checkout. Agentic commerce needs both, joined in feeds or APIs. That makes [PIM integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/) a prerequisite: which system publishes each field, how often and through which channel connector.

PIM vendors are adding agent access. Stibo Systems, inriver and Bluestone PIM advertise MCP servers that let AI agents query governed product data, and Akeneo publishes guidance on preparing product data for agentic commerce. For brands selling through retailers, [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/) remains the main route, as retailers feed the agents.

## What should teams do now?

Start with an audit of the top revenue SKUs against the OpenAI and Google feed specifications. Measure completeness of identifiers, core attributes and variant data, then close gaps at the source in PIM. Assign an owner for each agent channel, decide which products are eligible for agent checkout and monitor how products appear in assistant answers.

## Key takeaways

- AI shopping agents discover and buy through structured feeds and open protocols, chiefly ACP, UCP, AP2 and MCP.
- Protocols and programmes changed several times between September 2025 and May 2026, so build channel-neutral product data.
- Agents need identifiers, typed attributes, variant structure, current price and availability, and clear policies.
- PIM supplies the governed record, while commerce and ERP systems supply live price and stock.
- Audit top SKUs against published feed specifications and fix gaps at the source.

Q: What is the Agentic Commerce Protocol?
A: The Agentic Commerce Protocol (ACP) is an open standard co-developed by OpenAI and Stripe, announced on 29 September 2025 with Instant Checkout in ChatGPT. It defines how merchants share product feeds and handle checkout and delegated payment with AI agents. It is published under the Apache 2.0 licence, with the latest stable release dated 17 April 2026.
Q: What is Google's Universal Commerce Protocol?
A: The Universal Commerce Protocol (UCP) is an open-source standard announced by Google on 11 January 2026, developed with Shopify, Etsy, Wayfair, Target and Walmart. It covers the commerce journey from discovery to order management, works with the Agent Payments Protocol and supports REST, MCP and A2A connections. Google's implementation requires a Merchant Center account.
Q: Do B2B manufacturers need to prepare for agentic commerce?
A: Yes, in a different form. B2B buyers and procurement systems increasingly use AI assistants to search catalogues and compare specifications. Manufacturers benefit from the same groundwork: verified identifiers, classified products, complete technical attributes and documents, and an API or MCP server through which authorised agents can query governed product data.

## Article: AI auto-tagging and visual search in DAM
URL: https://pimatlas.com/knowledge/ai-auto-tagging-dam/
Category: AI and search | Updated: 2026-10-17

Summary: AI auto-tagging uses image recognition models to add keywords, objects, colours and text to digital assets automatically, while visual search finds assets by image or natural-language description using embeddings. Both cut manual tagging, but need a controlled vocabulary, confidence thresholds and review, and AI-generated images must carry provenance metadata such as IPTC Digital Source Type.

AI auto-tagging in a DAM (digital asset management) system uses image recognition to add descriptive metadata to assets automatically, and visual search uses vector embeddings to find assets by visual similarity or a plain-language description. Both reduce manual tagging effort. Reliable results need a controlled vocabulary, confidence thresholds, human review and preserved provenance metadata.

## How does AI auto-tagging work?

An auto-tagging service passes each uploaded image or video frame through a computer vision model, which returns labels with a confidence score. Google Cloud Vision label detection, for example, returns a description, a score from 0 to 1 and a topicality value per label, with 10 labels by default. Other models detect objects, faces, logos, dominant colours and embedded text through optical character recognition (OCR).

The DAM then maps those labels into metadata fields. Without mapping, generic labels flood the keyword field and reduce search precision. With mapping to a [DAM metadata taxonomy](https://pimatlas.com/knowledge/dam-metadata-taxonomy/), labels become controlled values such as "room type: kitchen" or "shot type: packshot".

## How does visual and semantic search work?

Visual search relies on embeddings: numerical vectors that represent the content of an image. Models trained on image and text pairs, such as CLIP (Radford and colleagues, 2021, trained on 400 million image-text pairs), place images and phrases in the same vector space. A search for "white sofa in a bright living room" then returns matching images even when no one tagged them that way.

This supports three common tasks:

- **Similarity search:** find all images that resemble a selected asset, useful for duplicate detection.
- **Natural-language search:** find assets by description without knowing the tag vocabulary.
- **Reverse image search:** match a supplier or retailer image to the master asset.

Embeddings are computed once per asset and stored in a vector index, so search cost is low after ingestion. Re-indexing is needed when the vendor changes the underlying model, which can shift result rankings.

## Which DAM AI features are vendors shipping?

Vendors name these features differently, and many bundle them with generative editing. Syndigo's AI GoPilots include image analysis that suggests tags, descriptions and metadata. Bluestone PIM lists AI DAM among its AI capabilities. Pimcore offers Pimcore Copilot for image manipulation and generation and AI-based upscaling. 4ALLPORTAL promotes Natural Language Search and facial recognition through a Microsoft Azure integration. THRON markets its AMBRA AI agent for product descriptions, translation and image editing.

When comparing, separate four capability types: tagging, search, generation and editing. Each has different accuracy, cost and compliance profiles. See [what is DAM](https://pimatlas.com/knowledge/what-is-dam/) for the wider feature set.

## How do you control auto-tagging quality?

| Control | Purpose | Practical setting |
|---|---|---|
| Controlled vocabulary | Keeps tags consistent and searchable | Map model labels to taxonomy values, discard the rest |
| Confidence threshold | Filters weak labels | Set per tag type, stricter for people and brands |
| Review queue | Catches errors before publishing | Route low-confidence and sensitive tags to librarians |
| Source flag | Separates AI from human tags | Store tag origin and model version |
| Accuracy sample | Measures precision over time | Label 200 to 500 assets, re-test after model changes |
| PIM link | Adds product identity | Inherit SKU, product line and category from PIM |

Product identity is the main limit of generic models: they recognise "running shoe" but not the specific model and colourway. That information comes from PIM through [DAM and PIM integration](https://pimatlas.com/knowledge/dam-pim-integration/), using filenames, GTINs or asset links.

## How should AI-generated images be handled?

Generative tools now create backgrounds, lifestyle scenes and variants in the DAM. These assets need provenance metadata. The IPTC Digital Source Type vocabulary defines trainedAlgorithmicMedia for content created with generative AI and compositeWithTrainedAlgorithmicMedia for content edited with it. Google Merchant Center requires AI-generated product images to carry this metadata and asks merchants not to remove it.

The EU AI Act (Regulation (EU) 2024/1689) adds Article 50 transparency duties from 2 August 2026, including machine-readable marking by providers of generative systems. Configure the DAM so that renditions and exports keep embedded metadata, and align image edits with [product image standards](https://pimatlas.com/knowledge/product-image-standards/) to avoid misrepresenting the product.

## Key takeaways

- Auto-tagging adds labels with confidence scores, and visual search finds assets through embeddings.
- Map AI labels to a controlled vocabulary and set confidence thresholds per tag type.
- Product identity comes from PIM, so link assets to SKUs and GTINs.
- Store tag origin and model version, and re-test accuracy after model changes.
- Preserve IPTC Digital Source Type metadata on AI-generated images through every export.

Q: How accurate is AI auto-tagging in a DAM?
A: Generic labels such as 'chair', 'outdoor' or 'blue' are usually reliable, while brand-specific concepts such as product lines, campaign names or SKU identity need custom models or PIM data. Recognition services return confidence scores, so set thresholds per tag type, review low-confidence tags and measure precision on a labelled sample of your own assets.
Q: What is the difference between auto-tagging and visual search?
A: Auto-tagging writes keywords into asset metadata, which users then search with text. Visual search compares assets by their visual content using embeddings, so users can find similar images or search with a natural-language phrase without matching tags. Many DAM systems combine both, using tags for governance and filters and embeddings for discovery.
Q: Do AI-generated product images need to be labelled?
A: Google Merchant Center requires AI-generated images to carry IPTC Digital Source Type metadata such as TrainedAlgorithmicMedia and asks merchants not to strip it. Under EU AI Act Article 50, providers of generative AI systems must mark outputs in a machine-readable format, and deployers must disclose deep fakes. Keep provenance metadata intact through the DAM.

## Article: The EU AI Act: what product content teams need to know
URL: https://pimatlas.com/knowledge/eu-ai-act-product-content/
Category: AI and search | Updated: 2026-10-17

Summary: For product content teams, the EU AI Act (Regulation (EU) 2024/1689) matters mainly through Article 50 transparency duties, applicable from 2 August 2026: disclosing AI chatbots, machine-readable marking of synthetic content and labelling deep fakes. Product copy and enrichment rarely count as high-risk, whose Annex III obligations the Digital Omnibus moved to 2 December 2027.

For product content teams, the EU AI Act (Regulation (EU) 2024/1689) applies mainly through the Article 50 transparency duties that took effect on 2 August 2026. These cover AI chatbots, machine-readable marking of AI-generated content and disclosure of deep fakes. Typical PIM and DAM uses of AI are not high-risk, and the Digital Omnibus has postponed high-risk obligations in any case.

## What is the EU AI Act timeline after the Digital Omnibus?

The AI Act entered into force on 1 August 2024 and applies in stages. The Digital Omnibus on AI, Regulation (EU) 2026/1744 of 8 July 2026, entered into force on 27 July 2026 and moved the high-risk deadlines.

| Date | What applies |
|---|---|
| 2 February 2025 | Prohibited practices and AI literacy (Article 4) |
| 2 August 2025 | Obligations for general-purpose AI models |
| 2 August 2026 | General application, including Article 50 transparency duties |
| 2 December 2026 | End of Article 50(2) marking grace period for generative systems placed on the market before 2 August 2026 |
| 2 December 2027 | High-risk systems listed in Annex III |
| 2 August 2028 | High-risk AI embedded in products under Annex I legislation |

The Commission proposed the Omnibus on 19 November 2025 and political agreement followed on 7 May 2026. Article 50 itself was not postponed.

## Are product content teams providers or deployers?

The role decides the duty. A provider develops an AI system, or has one developed, and places it on the market or puts it into service under its own name or trademark (Article 3(3)). A deployer uses an AI system under its authority in a professional context (Article 3(4)).

A brand that uses a PIM vendor's generative feature is normally a deployer. The PIM or model vendor is the provider. A retailer that builds its own shopping assistant and offers it under its own brand may be a provider of that assistant.

## What does Article 50 require in practice?

Article 50 sets four duties relevant to product content:

1. **Chatbot disclosure (Article 50(1), providers).** AI systems that interact directly with people, such as shopping assistants and support bots, must inform users they are dealing with AI unless it is obvious.
2. **Machine-readable marking (Article 50(2), providers).** Outputs of systems generating synthetic audio, images, video or text must be marked and detectable as AI-generated. Assistive standard editing that does not substantially alter the input is excluded.
3. **Deep fakes (Article 50(4), deployers).** Image, audio or video that resembles existing persons, objects, places or events and would falsely appear genuine must be disclosed as AI-generated or manipulated.
4. **Public-interest text (Article 50(4), deployers).** AI-generated text published to inform the public on matters of public interest must be disclosed, unless it underwent human review or editorial control.

Disclosure must be clear and distinguishable and given at the latest at the first interaction or exposure. Breaches of Article 50 can lead to fines of up to EUR 15 million or 3% of worldwide annual turnover, whichever is higher, with the lower figure applying to SMEs.

## Where does this touch PIM and DAM work?

Most product descriptions, translations and attribute enrichment fall outside the public-interest text duty. Risk concentrates in imagery and customer-facing assistants. A photorealistic AI scene showing a product in a real, recognisable location, or a generated image that misrepresents the product, can approach the deep fake definition. Our guide to [AI auto-tagging and visual search in DAM](https://pimatlas.com/knowledge/ai-auto-tagging-dam/) covers provenance metadata such as IPTC Digital Source Type.

The Commission published a Code of Practice on marking and labelling AI-generated content on 10 June 2026 and guidelines on Article 50 transparency obligations. Signing the Code is voluntary. Channel rules apply in parallel: Google Merchant Center asks for AI-generated titles and descriptions in separate structured attributes, as described in [generative AI for product descriptions](https://pimatlas.com/knowledge/generative-ai-product-descriptions/).

## What should product content teams do now?

- **Inventory AI use.** List every AI feature in PIM, DAM, syndication and site search, with provider, deployer role and output type.
- **Check provider marking.** Ask vendors how generated images and text are marked and whether exports keep the marks.
- **Label assistants.** Add clear AI notices to chatbots and shopping assistants.
- **Govern imagery.** Define rules for generated scenes, models and locations, with review for deep fake risk.
- **Record provenance.** Store origin, model and reviewer for AI-generated values within your [product data governance](https://pimatlas.com/knowledge/product-data-governance/) framework.
- **Train staff.** Cover AI literacy under Article 4, as amended by the Omnibus.

AI rules sit alongside product law. Safety information under the [GPSR](https://pimatlas.com/knowledge/gpsr-product-information/) must stay accurate whichever tool produced it.

## Key takeaways

- Article 50 transparency duties apply from 2 August 2026, with a marking grace period to 2 December 2026 for older generative systems.
- The Digital Omnibus, Regulation (EU) 2026/1744, moved Annex III high-risk obligations to 2 December 2027 and Annex I to 2 August 2028.
- Product copy is generally outside the public-interest text duty, while chatbots and synthetic imagery need attention.
- Brands using vendor AI features are usually deployers, and vendors are providers.
- Inventory AI use, label assistants, preserve provenance and govern generated imagery.

Q: Does the EU AI Act require labelling AI-written product descriptions?
A: Generally not. Article 50(4) requires deployers to disclose AI-generated text published to inform the public on matters of public interest, unless it underwent human review or editorial control with editorial responsibility. Ordinary product copy is outside that scope. Providers of generative systems must still mark outputs in a machine-readable format, and channels such as Google Merchant Center have their own labelling rules.
Q: When do the AI Act transparency obligations apply?
A: Article 50 obligations apply from 2 August 2026. The Digital Omnibus on AI, Regulation (EU) 2026/1744, gives providers of generative AI systems placed on the market before 2 August 2026 until 2 December 2026 to meet the machine-readable marking duty in Article 50(2). Other Article 50 duties were not postponed.
Q: Is a shopping chatbot on our website covered by the AI Act?
A: Yes. Article 50(1) requires AI systems that interact directly with people to be designed so users are informed they are dealing with AI, unless it is obvious. A retailer or brand that puts its own assistant into service under its own name may be the provider. Show a clear notice at the first interaction at the latest.

## Article: Product data syndication: reaching retailers and marketplaces
URL: https://pimatlas.com/knowledge/product-data-syndication/
Category: Standards and regulation | Updated: 2026-10-17

Summary: Product data syndication is the controlled distribution of product information from a PIM or source system to retailers, marketplaces, distributors and data pools, transformed to each recipient's format, taxonomy and validation rules. It relies on four routes: GS1 GDSN, B2B catalogue standards such as BMEcat, channel APIs and feed files, and retailer supplier portals.

Product data syndication is the process of publishing governed product information from a PIM (product information management) system to every external party that sells or specifies the product: retailers, marketplaces, distributors, data pools and comparison sites. Each recipient has its own format, taxonomy, mandatory attributes and validation rules, so syndication is mostly a mapping, conversion and feedback problem. Organisations that treat it as a one-off export usually see rejected listings, delayed launches and inconsistent content across channels.

## What does product data syndication involve?

Syndication sits at the outbound end of the [product content supply chain](https://pimatlas.com/knowledge/product-content-supply-chain/). A syndication capability typically performs five tasks:

1. **Mapping** internal attributes, units and code lists to each recipient's data model.
2. **Classifying** products into the recipient's taxonomy (a retailer category tree, GS1 GPC, ETIM or ECLASS).
3. **Transforming** values: unit conversion, text length limits, image renditions, language variants.
4. **Validating** against the recipient's rules before sending, so errors surface internally.
5. **Processing feedback**: acknowledgements, rejections and retailer change requests returned to the data owner.

The quality of the source data limits everything downstream. A complete, well-modelled PIM record with consistent units makes mapping repeatable; a weak record forces manual fixes per channel. [Product data quality](https://pimatlas.com/knowledge/product-data-quality/) metrics should therefore be measured per channel, since completeness against one retailer's template says little about another's.

## Which syndication routes exist?

Most brands and manufacturers use several routes in parallel. The right mix depends on the sector and on how recipients prefer to receive data.

| Route | Typical recipients | Format and mechanism | Key standard or specification |
|---|---|---|---|
| GS1 GDSN | Grocery, health and beauty, DIY and healthcare retailers | Certified data pool, publish and subscribe | GS1 Global Data Model, GTIN |
| B2B catalogue files | Wholesalers, distributors, procurement platforms | XML or JSON files per catalogue | BMEcat, ETIM xChange, ECLASS |
| Marketplace and retailer APIs | Amazon, Zalando, bol and similar platforms | API calls or scheduled feeds | Channel-specific category templates |
| Feed files | Shopping engines, affiliate and comparison sites | CSV, XML or JSON feeds | Channel-specific feed specifications |
| Supplier portals | Retailers without automated intake | Manual upload or spreadsheet templates | Retailer-defined templates |

GDSN (Global Data Synchronisation Network) suits retailers that have standardised on GS1. A supplier joins a certified data pool, loads and registers its items, and publishes them to subscribing trading partners, who confirm receipt. The [GS1, GTIN and GDSN guide](https://pimatlas.com/knowledge/gs1-gtin-gdsn/) covers the mechanics. For technical products, distributors in electrical, installation and HVAC sectors expect ETIM-classified data in BMEcat or ETIM xChange; ETIM International released ETIM xChange version 2.0 on 27 November 2025. The [ETIM, BMEcat and ECLASS article](https://pimatlas.com/knowledge/etim-bmecat-eclass/) explains these formats.

## Should syndication run inside the PIM or in a separate platform?

Both models are common. PIM-native syndication keeps mapping logic close to the data model and avoids another licence. Dedicated syndication or feed management platforms, such as [Productsup](https://pimatlas.com/vendors/productsup/), and suites that combine PIM with retailer networks, such as Salsify and Syndigo, maintain channel connectors and templates as recipients change their specifications.

Evaluation criteria that separate the options:

- Number and depth of maintained channel connectors relevant to the buyer's markets.
- Speed at which connector templates are updated after a channel changes its requirements.
- Support for GDSN certification, BMEcat and ETIM xChange export.
- Channel-level completeness scoring and pre-send validation.
- Rejection handling: whether errors return to the PIM user who owns the attribute.
- Asset handling: renditions, naming rules and image requirements per channel.

The [integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/) decision also matters: syndication volumes, delta updates and scheduling should be designed alongside ERP and commerce flows.

## How does regulation affect syndicated content?

Regulation increasingly defines mandatory content fields. Under the General Product Safety Regulation (Regulation (EU) 2023/988), which applies from 13 December 2024, online offers must show manufacturer details, product identification including a picture, and warnings or safety information. Food sold online must carry the mandatory particulars of Regulation (EU) No 1169/2011 before purchase. These fields must flow through syndication to every channel where consumers see the offer, so legal attributes need the same mapping and validation discipline as marketing attributes.

## Key takeaways

- Syndication maps, transforms, validates and delivers PIM data to each external recipient's specification, and processes feedback.
- Most organisations combine GDSN, B2B catalogue standards, marketplace APIs, feed files and portals.
- Channel-level completeness and pre-send validation reduce rejected listings and launch delays.
- Choose between PIM-native and dedicated syndication by connector coverage, update speed and error handling.
- Regulatory fields under GPSR and food information rules must be syndicated with the same rigour as marketing content.

Q: What is the difference between PIM and product data syndication?
A: A PIM system creates, enriches and governs the master version of product information. Syndication takes that governed record and delivers it to external recipients, mapping attributes, units, taxonomies and assets to each channel's specification and handling validation feedback. Some PIM suites include syndication modules; other organisations add a dedicated feed management or syndication platform alongside the PIM.
Q: Do I need GDSN to sell through European grocery retailers?
A: Many grocery, health and beauty and DIY retailers that use GS1 standards ask suppliers to publish product data through a certified GDSN data pool. Requirements differ by retailer and country, so suppliers should confirm each retailer's onboarding rules, required attributes and target market with the retailer or the local GS1 member organisation before choosing a data pool.
Q: How many attributes does a marketplace listing typically need?
A: There is no single number. Each marketplace defines category-specific attribute sets, with mandatory, conditional and recommended fields that change over time. A practical approach is to map each channel's category templates against the PIM data model, measure completeness per channel and track rejected listings, so gaps are visible before products go live.

## Article: GS1, GTINs and GDSN explained
URL: https://pimatlas.com/knowledge/gs1-gtin-gdsn/
Category: Standards and regulation | Updated: 2026-10-17

Summary: GS1 is the not-for-profit standards organisation behind the barcode. A GTIN (Global Trade Item Number) uniquely identifies each trade item, GDSN (Global Data Synchronisation Network) shares the attributes behind that GTIN between trading partners through certified data pools, and GS1 Digital Link turns the identifier into a web address for QR codes.

GS1 standards give every product a globally unique identity and a shared language for describing it. The GTIN (Global Trade Item Number) identifies the trade item, GDSN (Global Data Synchronisation Network) shares the master data attached to that GTIN between suppliers and retailers, and GS1 Digital Link encodes the GTIN as a web address so one QR code can serve checkout scanners and consumers. For a PIM (product information management) team, the GTIN is the key that connects internal records to every external system.

## What is a GTIN and how is it structured?

A GTIN is a number that identifies any trade item that can be priced, ordered or invoiced at any point in the supply chain. The brand owner assigns it, using capacity licensed from a GS1 member organisation, and GS1 GTIN Management Rules govern when a change to a product requires a new number.

| Format | Length | Typical use |
|---|---|---|
| GTIN-8 | 8 digits | Very small consumer packs (EAN-8 barcode) |
| GTIN-12 | 12 digits | Consumer units in North America (UPC-A barcode) |
| GTIN-13 | 13 digits | Consumer units in most other markets (EAN-13 barcode) |
| GTIN-14 | 14 digits | Cases and other non-retail trade units |

In the [product data model](https://pimatlas.com/knowledge/product-data-modelling/), the GTIN belongs at the level that is actually traded: each variant, pack size and case level needs its own number. Storing GTINs as text with leading zeros preserved, and validating the check digit on entry, prevents common integration errors.

## How does GDSN work?

GDSN is a network of certified data pools that synchronise product master data between trading partners. GS1 describes five steps: the supplier loads its item data into a data pool, registers the items, the recipient subscribes, the supplier publishes the data and the recipient confirms receipt. The content follows the GS1 Global Data Model, a globally consistent set of core attributes needed to list, store, move and sell products, extended by local market requirements.

For suppliers, GDSN replaces many retailer-specific spreadsheets with one publication route. The practical work sits in the PIM: modelling the packaging hierarchy, holding logistics attributes such as dimensions and weights per level, and mapping attributes to the target market's code lists. GDSN is one route within broader [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/), and many grocery, health and beauty and DIY retailers require it.

## What are GS1 Digital Link and Sunrise 2027?

GS1 Digital Link is the GS1 standard for expressing identifiers such as GTINs, GLNs and SSCCs as web URIs. GS1 lists URI syntax version 1.5.0 (June 2024) and a conformant resolver standard version 1.0.0 (February 2024). A QR code carrying a Digital Link URI can be scanned at the till to extract the GTIN, and by a consumer's phone to reach online information; the destination can be changed online without reprinting packaging.

Sunrise 2027 is the GS1 industry ambition that, by the end of 2027, QR codes with GS1 standards are widely adopted, with retailers able to scan them at point of sale. The GS1 global endorsement statement notes that point-of-sale systems must read both 1D and 2D barcodes during the transition, and GS1 US advises retailers that the first step is installing scanners that can read 2D codes and extract the GTIN. Both QR codes and GS1 DataMatrix can carry a GTIN with attributes such as expiry date or serial number.

## Why does this matter for PIM and regulation?

Three developments raise the importance of clean GS1 data in the PIM:

- **2D barcodes and batch data**: encoding expiry dates or serial numbers requires the PIM and packaging workflow to manage more than the GTIN.
- **Consumer information via QR**: a Digital Link destination needs product content, such as ingredients, allergens or instructions, that is complete and current. Online food content must still meet [Regulation (EU) No 1169/2011](https://pimatlas.com/knowledge/food-information-regulation-1169/).
- **Digital Product Passport**: product identifiers and data carriers are central to the EU [Digital Product Passport](https://pimatlas.com/knowledge/digital-product-passport/) under Regulation (EU) 2024/1781, and GTINs are one of the identifier schemes discussed in its standardisation work.

Vendors such as Syndigo and Salsify operate GDSN-certified services; buyers should confirm certification status and supported target markets directly with each provider.

## Key takeaways

- The GTIN identifies each traded item and packaging level; the brand owner assigns it under GS1 rules.
- GDSN synchronises master data through certified data pools using the GS1 Global Data Model.
- GS1 Digital Link turns identifiers into web URIs so one QR code can serve checkout and consumers.
- Sunrise 2027 targets wide adoption of QR codes with GS1 standards at retail point of sale by the end of 2027.
- The PIM should own GTIN allocation, packaging hierarchies and the content that 2D codes link to.

Q: Who assigns a GTIN to a product?
A: The brand owner, meaning the organisation that owns the product specification, normally assigns the GTIN, wherever the item is made. It does so from a numbering capacity licensed by its local GS1 member organisation. GS1 GTIN Management Rules define when a product change requires a new GTIN, so allocation should be governed centrally, often in the PIM or ERP.
Q: Is the EAN barcode being replaced by QR codes in 2027?
A: GS1's Sunrise 2027 ambition is that QR codes and other 2D barcodes carrying GS1 identifiers are widely adopted by the end of 2027, with retail point-of-sale systems able to read both 1D and 2D codes. The transition is gradual, so linear barcodes continue to be scanned while brands add 2D codes to packaging.
Q: What is a GDSN data pool?
A: A GDSN data pool is a service certified against GS1 standards that stores a supplier's product data and synchronises it with subscribing trading partners. Suppliers load and register their items in their chosen pool, publish them to subscribing recipients and receive confirmations. One publication can then reach all subscribing trading partners, which reduces the number of retailer-specific templates a supplier maintains.

## Article: ETIM, BMEcat and ECLASS: B2B product data standards
URL: https://pimatlas.com/knowledge/etim-bmecat-eclass/
Category: Standards and regulation | Updated: 2026-10-17

Summary: ETIM and ECLASS are classification standards that define product classes and the technical features or properties that describe them. BMEcat and ETIM xChange are exchange formats that carry catalogue data, including those classifications, between manufacturers, wholesalers and buyers. Most technical B2B suppliers combine one classification with one exchange format.

ETIM, ECLASS and BMEcat solve two different problems in B2B product data. ETIM and ECLASS are classification standards: they define product classes and the features or properties that describe each class. BMEcat, and the newer ETIM xChange, are exchange formats: they package catalogue data, including those classifications, for transfer between manufacturers, wholesalers and procurement systems. A supplier of technical products usually needs one classification and one exchange format per customer group.

## What is the difference between ETIM, ECLASS and BMEcat?

| Standard | Type | Maintained by | Current version (October 2026) | Typical sectors |
|---|---|---|---|---|
| ETIM | Classification model | ETIM International and national ETIM organisations | ETIM 10.0 (released December 2024); ETIM 11.0 planned for 1 December 2026 | Electrical, installation, HVAC, plumbing, building technology |
| ECLASS | Classification and property dictionary | ECLASS e.V. | ECLASS 16.0 (released 28 November 2025); annual release cycle | Industrial procurement, automation, engineering, MRO |
| BMEcat | XML catalogue exchange format | Originated with BME, Fraunhofer IAO and university partners | BMEcat 2005 family, with ETIM's BMEcat 2005 guideline version 5.0 | Wholesale, e-procurement, distributor catalogues |
| ETIM xChange | JSON exchange format | ETIM International | Version 2.0 (released 27 November 2025) | ETIM-based sectors moving from file formats to JSON |

## How does ETIM classify products?

ETIM describes itself as the international classification standard for technical products. Each ETIM class, such as a circuit breaker or a pipe fitting, carries a defined list of features with fixed units and value lists. A supplier classifies each product into one class and fills in the features, which lets a wholesaler compare products from different manufacturers on the same technical basis.

ETIM moved to a release cycle of roughly two years. ETIM International's release planning for ETIM 11.0 set a change request deadline of 30 June 2026, processing of open requests before 15 October 2026, a beta of about one month, and a planned final release on 1 December 2026. Suppliers should plan a mapping update after each release, since classes and features can be added, split or deprecated. ETIM also offers ETIM MC (Modelling Classes) for configurable products.

## What does ECLASS add?

ECLASS is a hierarchical classification combined with a property dictionary, built with unique identifiers for classes, properties and values. ECLASS e.V. publishes a release every year. Release 16.0, published on 28 November 2025, added 995 classes, 1,253 properties, 985 values and 126 value lists compared with Release 15.0, and is fully translated into 29 languages. It is distributed in two representations: BASIC (CSV or XML, with a reduced property set) and ADVANCED (XML only, with property blocks, polymorphism and cardinality). Release 16.0 also introduced a first draft of recycling information, which ECLASS states will be developed further in Release 17, a sign of the standard's growing role in the [Digital Product Passport](https://pimatlas.com/knowledge/digital-product-passport/).

## Where do BMEcat and ETIM xChange fit?

BMEcat is an XML format for electronic catalogues. Version 1.0 was approved in November 1999, and the BMEcat 2005 family remains in wide use. BMEcat can carry ETIM or ECLASS classifications, prices, media references and product relationships. ETIM International publishes a guideline that defines how ETIM data is expressed in BMEcat 2005; version 5.0 was published on 30 June 2021.

ETIM xChange is ETIM International's JSON-based format, designed for API-style exchange and validation. Version 2.0 was officially released on 27 November 2025. In practice, wholesalers in some countries accept both, so suppliers export each format from the same PIM data, as part of their wider [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/).

## What should a PIM support?

When evaluating PIM systems for technical B2B products, the following capabilities separate strong fits:

- Import of ETIM and ECLASS releases, with change reports between versions.
- Mapping from an internal attribute model to several classifications without duplicate data entry.
- Unit and value list handling that matches ETIM and ECLASS codes.
- Validated export to BMEcat 2005 under the ETIM guideline and to ETIM xChange.
- Support for multiple classification versions in parallel while customers migrate.

These requirements belong in the [taxonomy design](https://pimatlas.com/knowledge/product-taxonomy-design/) and in RFPs for [wholesale and distribution](https://pimatlas.com/knowledge/pim-for-wholesale-distribution/) projects.

## Key takeaways

- ETIM and ECLASS classify products and define features; BMEcat and ETIM xChange carry the data.
- ETIM 10.0 is current; ETIM 11.0 is planned for 1 December 2026.
- ECLASS 16.0 was released on 28 November 2025, and ECLASS publishes annually.
- ETIM xChange 2.0 was released on 27 November 2025; BMEcat 2005 remains widely used.
- A PIM should map one internal model to several classifications and export each format.

Q: Should a manufacturer choose ETIM or ECLASS?
A: The choice usually follows the customers. Electrical, installation, HVAC and plumbing wholesalers in Europe widely request ETIM data, while ECLASS is common in industrial procurement, automation and engineering contexts. Many manufacturers maintain both, holding one internal attribute model in the PIM and mapping it to each classification so features are entered once.
Q: What is the current ETIM version?
A: ETIM 10.0 was released in December 2024 and is the current version as of October 2026. ETIM International has planned the final release of ETIM 11.0 for 1 December 2026, preceded by a beta period of about one month. Suppliers should confirm with their local ETIM organisation which versions their customers accept.
Q: Is BMEcat being replaced by ETIM xChange?
A: ETIM xChange is a newer JSON-based exchange format from ETIM International, with version 2.0 released on 27 November 2025. BMEcat remains widely used, including through ETIM's BMEcat 2005 guideline. Adoption of xChange varies by country and sector, so suppliers typically support both formats while their trading partners migrate.

## Article: The Digital Product Passport and PIM: preparing for ESPR
URL: https://pimatlas.com/knowledge/digital-product-passport/
Category: Standards and regulation | Updated: 2026-10-17

Summary: The Digital Product Passport (DPP) is a product-level digital record required under the Ecodesign for Sustainable Products Regulation (EU) 2024/1781. The EU DPP registry was set up in July 2026, but obligations apply only once product-specific delegated acts are adopted. Iron and steel come first, with textiles, tyres and aluminium indicated for 2027.

The Digital Product Passport (DPP) is a structured, machine-readable record attached to a product through a data carrier such as a QR code, created under the Ecodesign for Sustainable Products Regulation (ESPR), Regulation (EU) 2024/1781. ESPR entered into force on 18 July 2024, and the Commission set up the central DPP registry in July 2026. Obligations for companies apply only once a product-specific delegated act requires a passport. PIM (product information management) teams therefore have time to prepare, though the data work is substantial.

## What is the current status of the DPP in October 2026?

| Milestone | Date | Status |
|---|---|---|
| ESPR entry into force | 18 July 2024 | In force |
| ESPR and Energy Labelling Working Plan 2025-2030 adopted | 16 April 2025 | Adopted |
| Rules on destruction of unsold apparel and footwear (Delegated Regulation (EU) 2026/296, Implementing Regulation (EU) 2026/2) adopted | 9 February 2026 | Adopted |
| Destruction ban applies to large companies | 19 July 2026 | Applies |
| Deadline for the Commission to set up the DPP registry (Article 13) | 19 July 2026 | Registry launched |
| Implementing Regulation (EU) 2026/1778 on registry arrangements | Adopted 16 July 2026, in force 6 August 2026 | In force |
| Battery passport under Regulation (EU) 2023/1542 | 18 February 2027 | Upcoming, separate regime |

As of October 2026, no product-specific ESPR delegated act setting DPP requirements had been adopted. The act for iron and steel is the first expected.

The registry is an index. It stores the unique identifiers that link a product to its passport, while the passport data itself stays with the economic operator or a DPP service provider acting on its behalf. The registry also records commodity codes for goods entering free circulation, so customs authorities can use it to check that imported products in scope have a passport once a delegated act applies.

## Which product groups come first?

The working plan adopted on 16 April 2025 sets priorities and indicative adoption years for delegated acts:

- **Iron and steel**: 2026
- **Textiles, particularly apparel**: 2027
- **Tyres**: 2027
- **Aluminium**: 2027
- **Furniture**: 2028
- **Mattresses**: 2029
- **Horizontal repairability requirements**, including a repairability score for consumer electronics and small household appliances: 2027
- **Horizontal recycled content and recyclability requirements for electrical and electronic equipment**: 2029

These years refer to adoption. Requirements typically apply after a transition of at least 18 months, so the first passports for most consumer goods are unlikely before 2028 or later. Each delegated act will define the exact data points, access rights and data carrier for its product group.

## What data will a passport contain?

The Commission describes the DPP as a digital identity card for products, components and materials. Depending on the delegated act, it may include:

1. **Identification**: a unique product identifier, linked to operator and facility identifiers.
2. **Technical performance** and compliance documentation.
3. **Materials and origin**, including substances of concern.
4. **Use, repair and maintenance** information, including spare parts.
5. **End-of-life**: recycling and disassembly instructions.
6. **Environmental footprint** across the lifecycle.

Identifiers and data carriers are where existing standards connect. GTINs and [GS1 Digital Link](https://pimatlas.com/knowledge/gs1-gtin-gdsn/) are among the identifier schemes discussed in the European DPP standardisation work by CEN-CENELEC JTC 24, and ECLASS has added a first draft of recycling information to its classification, as described in the [ETIM, BMEcat and ECLASS guide](https://pimatlas.com/knowledge/etim-bmecat-eclass/).

## How should PIM teams prepare?

A PIM already holds a large share of passport-relevant data. The gaps tend to be traceability, material composition at component level, and lifecycle data owned outside marketing. Practical preparation steps:

- Inventory passport-relevant attributes per product group and name a source system for each.
- Introduce identifiers at the required granularity: model, batch or item, depending on the expected delegated act.
- Model materials, substances and supplier declarations as structured attributes with evidence documents.
- Extend [data governance](https://pimatlas.com/knowledge/product-data-governance/) to cover compliance owners, approval workflows and audit trails.
- Design the [integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/) so PLM, ERP and supplier data can be published to a DPP service with access control.

Organisations in iron and steel, textiles and tyres have the shortest runway and should follow the consultations on their delegated acts.

## Key takeaways

- ESPR (Regulation (EU) 2024/1781) entered into force on 18 July 2024; DPP obligations depend on product-specific delegated acts.
- The DPP registry was set up in July 2026, with arrangements in Implementing Regulation (EU) 2026/1778.
- Iron and steel is indicated for 2026; textiles, tyres and aluminium for 2027; furniture 2028; mattresses 2029.
- No product-specific DPP delegated act had been adopted as of October 2026.
- PIM teams should map attributes to source systems, strengthen identifiers and extend governance now.

Q: When does the Digital Product Passport become mandatory?
A: There is no single date. A DPP becomes mandatory for a product group only when a delegated act under Regulation (EU) 2024/1781 sets it, usually with a transition of at least 18 months. As of October 2026 no product-specific ESPR delegated act requiring a DPP had been adopted; iron and steel is the first group indicated, for 2026. The battery passport follows separate rules from 18 February 2027.
Q: Does a PIM system become the Digital Product Passport?
A: A PIM typically holds much of the commercial and technical data a passport needs, such as identifiers, materials, instructions and compliance documents. Sustainability, supply chain and lifecycle data often sit in PLM, ERP, LCA or supplier systems. Most architectures combine these sources and publish the passport through a DPP service that handles hosting, access rights and registry interfaces.
Q: What is the EU DPP registry?
A: The registry is a central EU system that Article 13 of Regulation (EU) 2024/1781 required the Commission to set up by 19 July 2026. It stores unique identifiers that link to passports, while the passport data itself remains with the economic operator or its service provider. Commission Implementing Regulation (EU) 2026/1778 of 16 July 2026 sets its implementation arrangements.

## Article: GPSR: product safety information obligations for online sellers
URL: https://pimatlas.com/knowledge/gpsr-product-information/
Category: Standards and regulation | Updated: 2026-10-17

Summary: The General Product Safety Regulation (EU) 2023/988 applies from 13 December 2024. Under Article 19, every online offer of a consumer product must clearly show the manufacturer's name and contact details, an EU responsible person where the manufacturer is outside the EU, product identification including a picture, and any warnings or safety information.

The General Product Safety Regulation (GPSR), Regulation (EU) 2023/988, applies from 13 December 2024 and sets minimum product information that every online offer of a consumer product in the EU must display. Article 19 lists four categories: manufacturer identity and contact details, an EU responsible person for manufacturers outside the EU, product identification with a picture, and warnings or safety information. For product data teams, these are mandatory attributes that must exist in the PIM (product information management) system and reach every channel.

This article describes the obligations factually; it does not constitute legal advice.

## What must an online listing show under GPSR Article 19?

Article 19 applies when economic operators make products available online or through other distance sales. The offer must clearly and visibly indicate at least the following:

| Article 19 point | Required information | Typical PIM attributes |
|---|---|---|
| (a) | Name, registered trade name or registered trade mark of the manufacturer, and the postal and electronic address where it can be contacted | Manufacturer entity, trade mark, postal address, email or web contact |
| (b) | Where the manufacturer is not established in the EU: name, postal and electronic address of the responsible person under Article 16(1) GPSR or Article 4(1) of Regulation (EU) 2019/1020 | EU responsible person entity and contact details |
| (c) | Information allowing identification of the product, including a picture, its type and any other product identifier | Main image, type or model, GTIN, batch or serial reference |
| (d) | Any warning or safety information to be affixed to the product or packaging or included in an accompanying document, in a language easily understood by consumers as determined by the Member State | Localised warnings, safety instructions, pictograms, documents |

Point (d) creates a language requirement per market. A product sold in Belgium, Germany and Poland needs safety information in the languages each Member State specifies, which makes localisation workflows a compliance matter.

## What do manufacturers and importers have to provide?

The online information derives from obligations on the physical product. Under Article 9, manufacturers must ensure their products carry a type, batch or serial number or another identifier that consumers can easily see; indicate their name, trade name or trade mark and postal and electronic address on the product, packaging or accompanying document; and supply clear instructions and safety information in a language determined by the Member State. Article 16 requires an economic operator established in the EU for products placed on the EU market, with its details on the product, packaging, parcel or accompanying document.

For PIM teams, this means the same data should drive packaging artwork, instructions and online offers, so a single governed record avoids discrepancies between label and listing.

## What changes for online marketplaces?

Article 22 places separate duties on online marketplace providers. They must register with the Safety Gate Portal and designate a single point of contact for market surveillance authorities, provide a contact point for consumers, act on authority orders to remove dangerous offers within two working days, and process product safety notices within three working days. Article 22(9) requires them to design their online interface so traders can supply the required product safety information.

In practice, marketplaces have added mandatory fields for manufacturer, responsible person and safety documents to their listing templates. Brands and retailers should map these fields in their [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/) set-up so the values come from the PIM.

## How should a PIM support GPSR compliance?

GPSR data is mostly master data about organisations and documents. The following capabilities reduce manual effort:

- **Reference entities** for manufacturers and responsible persons, linked to products, so an address change updates every affected offer.
- **Structured identifiers**: model or type, GTIN and batch logic in the [product data model](https://pimatlas.com/knowledge/product-data-modelling/).
- **Localised safety content** with completeness rules per target market language.
- **Document management** for instructions and safety sheets, with versions and validity dates.
- **Channel validation** that blocks publication when Article 19 fields are empty.
- **Governance**: named owners for compliance attributes and an audit trail, as part of [product data governance](https://pimatlas.com/knowledge/product-data-governance/).

Fashion and general merchandise retailers with large supplier bases face the heaviest load, since they must collect manufacturer and safety data from many suppliers before listing.

## Key takeaways

- GPSR, Regulation (EU) 2023/988, applies from 13 December 2024 and replaced Directive 2001/95/EC.
- Article 19 requires manufacturer details, an EU responsible person where relevant, product identification with a picture, and safety information in online offers.
- Safety information must be in a language determined by each Member State where the product is sold.
- Article 22 requires online marketplaces to enable traders to supply this information and to act on safety orders.
- Model manufacturers and responsible persons as linked entities in the PIM and validate Article 19 fields before publishing.

Q: When did the GPSR start to apply?
A: Regulation (EU) 2023/988 on general product safety was adopted on 10 May 2023, published in the Official Journal on 23 May 2023 and applies from 13 December 2024. It repealed the General Product Safety Directive 2001/95/EC and Council Directive 87/357/EEC. As an EU regulation, it applies directly in all Member States without national transposition.
Q: Does GPSR apply to B2B products?
A: GPSR covers products intended for consumers or likely to be used by consumers under reasonably foreseeable conditions, including products supplied in the course of a service. Products intended only for professional use fall outside its scope unless consumers are likely to use them. Specific Union harmonisation legislation takes precedence where it covers the same safety aspects.
Q: Who must provide GPSR information on an online marketplace?
A: The trader offering the product must display the Article 19 information in the offer. Online marketplace providers must design their interfaces so traders can provide it, register with the Safety Gate Portal and act on authority orders. Manufacturers and importers remain responsible for the underlying identification, contact and safety information on the product or its packaging.

## Article: Food information online: Regulation (EU) No 1169/2011 and PIM
URL: https://pimatlas.com/knowledge/food-information-regulation-1169/
Category: Standards and regulation | Updated: 2026-10-17

Summary: Regulation (EU) No 1169/2011 requires that food sold online shows the same mandatory particulars as the label, except the date of minimum durability or use by date, before the purchase is concluded and at no extra cost. That includes the name, ingredients with the 14 allergens emphasised, net quantity, nutrition declaration and food business operator.

Regulation (EU) No 1169/2011 on the provision of food information to consumers, dated 25 October 2011, requires that prepacked food sold online carries the same mandatory information as its label before the purchase is concluded. The only exception is the date of minimum durability or use by date, which must be available at delivery. The information must be free of supplementary charges to the consumer. For grocery retailers and food brands, this turns label content into regulated PIM (product information management) data.

This article describes the obligations factually; it does not constitute legal advice.

## Which particulars are mandatory under Article 9?

Article 9(1) lists twelve mandatory particulars. Article 14 applies them to distance selling.

| Article 9(1) | Particular | Online before purchase |
|---|---|---|
| (a) | Name of the food | Yes |
| (b) | List of ingredients | Yes |
| (c) | Ingredients or processing aids from Annex II (allergens) still present in the finished product | Yes |
| (d) | Quantity of certain ingredients or categories of ingredients | Yes |
| (e) | Net quantity of the food | Yes |
| (f) | Date of minimum durability or use by date | No, at delivery |
| (g) | Special storage conditions and/or conditions of use | Yes |
| (h) | Name or business name and address of the food business operator under Article 8(1) | Yes |
| (i) | Country of origin or place of provenance where required by Article 26 | Yes |
| (j) | Instructions for use where needed for appropriate use | Yes |
| (k) | Actual alcoholic strength by volume for beverages above 1.2% alcohol | Yes |
| (l) | Nutrition declaration | Yes |

The regulation has applied since 13 December 2014; the nutrition declaration became mandatory from 13 December 2016. Under Article 30(1), the declaration covers energy, fat, saturates, carbohydrate, sugars, protein and salt.

## What does Article 14 require for distance selling?

Article 14(1) states that for prepacked food offered by means of distance communication, mandatory particulars other than the durability date must be available before the purchase is concluded, appear on the material supporting the distance sale or be provided by other appropriate means clearly identified by the operator, and be provided without supplementary costs. All mandatory particulars must be available at delivery. Article 14(2) extends the obligation to non-prepacked food sold at a distance, for which allergen information under Article 44 is mandatory.

Article 8(1) assigns responsibility to the operator under whose name the food is marketed, or to the importer where that operator is outside the EU. Retailers selling third-party brands therefore depend on supplier data, which is why many grocery retailers request food information through [GS1 GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/).

## How must the 14 allergens be handled?

Annex II lists 14 substances or products causing allergies or intolerances:

1. Cereals containing gluten
2. Crustaceans
3. Eggs
4. Fish
5. Peanuts
6. Soybeans
7. Milk
8. Nuts
9. Celery
10. Mustard
11. Sesame seeds
12. Sulphur dioxide and sulphites
13. Lupin
14. Molluscs

Article 21(1)(b) requires these to be emphasised in the list of ingredients through a typeset that clearly distinguishes them, such as font, style or background colour. Online, the ingredient text needs markup that preserves this emphasis, and structured allergen attributes allow filtering and consistent display.

## How should a PIM model food information?

Food data combines free text, controlled values and numbers that must stay consistent with packaging. Requirements to test during selection:

- **Ingredient lists as rich text** with allergen emphasis that survives syndication to each channel.
- **Structured allergen attributes**: contains, may contain, free from, mapped to the 14 Annex II categories.
- **Nutrition tables** per 100 g or 100 ml, with optional per-portion values and validation of units.
- **Versioning by recipe change**, so online content updates when packaging does, with effective dates.
- **Localisation** per market language, aligned with label text.
- **Completeness rules** that block publication when any Article 9 particular is missing, as part of [product data quality](https://pimatlas.com/knowledge/product-data-quality/) scoring.
- **Label images** that meet [product image standards](https://pimatlas.com/knowledge/product-image-standards/), as supporting evidence alongside structured fields.

GS1 notes that a QR code with GS1 Digital Link can direct consumers to online ingredients, allergen and nutrition information, which raises the importance of keeping that content complete and current. Syndication to retailers is covered in the [product data syndication guide](https://pimatlas.com/knowledge/product-data-syndication/).

## Key takeaways

- Online food offers must show all Article 9 mandatory particulars, except the durability date, before purchase and at no extra cost.
- All particulars, including the durability date, must be available at delivery.
- The 14 Annex II allergens must be emphasised in the ingredient list.
- The nutrition declaration covers energy, fat, saturates, carbohydrate, sugars, protein and salt.
- Model ingredients, allergens and nutrition as versioned, structured PIM data aligned with the physical label.

Q: Does the best before date have to appear on a webshop?
A: No. Article 14(1) of Regulation (EU) No 1169/2011 exempts the date of minimum durability and the use by date from the information that must be available before the purchase is concluded. All mandatory particulars, including the date, must be available at the moment of delivery, which in practice means on the label of the delivered product.
Q: What are the 14 allergens under EU law?
A: Annex II of Regulation (EU) No 1169/2011 lists cereals containing gluten, crustaceans, eggs, fish, peanuts, soybeans, milk, nuts, celery, mustard, sesame seeds, sulphur dioxide and sulphites, lupin and molluscs. When used as ingredients or processing aids and still present in the finished product, they must be emphasised in the list of ingredients, for example in bold type.
Q: Can a product image of the label replace structured food information online?
A: The regulation requires mandatory particulars to be available before purchase and to be legible, but does not prescribe a specific digital format. Many retailers require structured attributes as well as label images, since structured data supports search, filtering, allergen display and accessibility. Readable label images alone can be hard to verify and may not meet retailer data specifications.

## Article: The European Accessibility Act and product content
URL: https://pimatlas.com/knowledge/european-accessibility-act-product-content/
Category: Standards and regulation | Updated: 2026-10-17

Summary: The European Accessibility Act, Directive (EU) 2019/882, applies from 28 June 2025 and covers e-commerce services. Online shops must meet the four accessibility principles of perceivability, operability, understandability and robustness, and must provide information on the accessibility of the products and services they sell when the responsible economic operator supplies it. Product content therefore needs alt text, accessible documents and accessibility attributes.

The European Accessibility Act (EAA), Directive (EU) 2019/882, has applied since 28 June 2025 through national transposition laws, and e-commerce services are within its scope. For product content, the EAA creates two kinds of work: making product pages, images and documents accessible to people with disabilities, and displaying accessibility information about the products sold. Both depend on structured data and asset metadata held in PIM (product information management) and DAM (digital asset management) systems.

This article describes the obligations factually; it does not constitute legal advice.

## What does the EAA require from e-commerce services?

Article 2(2)(f) brings e-commerce services into scope, defined in Article 3(30) as services provided at a distance through websites and mobile device-based services, by electronic means, to conclude a consumer contract. Recital 43 confirms that the e-commerce obligations apply to the online sale of any product or service. Member States had to transpose the directive by 28 June 2022 and apply the measures from 28 June 2025 (Article 31).

Annex I sets the requirements most relevant to product content:

| Annex I reference | Requirement | Product content implication |
|---|---|---|
| Section III(c) | Websites and mobile applications accessible in a consistent and adequate way, applying the four principles | Alt text for product images, captions for video, accessible tables and documents |
| Section IV(g)(i) | Provide information concerning the accessibility of the products and services being sold, when provided by the responsible economic operator | Accessibility attributes in the PIM, displayed on product pages |
| Section IV(g)(ii) | Accessible identification, security and payment functionality | Checkout and account flows (outside PIM scope) |
| Section IV(g)(iii) | Identification methods, electronic signatures and payment services that apply the four principles | Checkout and identity providers |

The directive expresses the four principles as perceivability, operability, understandability and robustness. They are the same principles that structure WCAG, the Web Content Accessibility Guidelines, which is why accessibility audits of product pages usually test against WCAG success criteria. For product content teams, perceivability carries most weight: a product image without a text alternative, or a specification shown only as an image, cannot be perceived by a screen reader user.

## Who is exempt, and what transition applies?

Article 4(5) exempts microenterprises providing services from the accessibility requirements. A microenterprise employs fewer than 10 people and has annual turnover or a balance sheet total not exceeding EUR 2 million. Article 32 lets Member States set a transitional period ending on 28 June 2030 for services provided with products lawfully used before 28 June 2025, and allows service contracts agreed before that date to run until expiry, for at most five years.

## What do manufacturers have to provide?

Products covered by the EAA, such as computers, smartphones, e-readers and self-service terminals, carry their own requirements. Annex I Section I requires information on the use of the product, including labelling, instructions and warnings, to be made available via more than one sensory channel, presented in fonts of adequate size and suitable shape, with sufficient contrast. Instructions must describe the product's accessibility functions, how to activate them and their compatibility with assistive technologies. This content usually originates in the PIM and must reach retailers through syndication, since retailers display it under Section IV(g)(i).

## How should PIM and DAM support accessible product content?

Accessibility becomes a set of attributes, asset fields and quality rules:

- **Alt text per image and per language**, stored in the DAM or PIM with the asset, following [product image standards](https://pimatlas.com/knowledge/product-image-standards/) and [DAM metadata design](https://pimatlas.com/knowledge/dam-metadata-taxonomy/).
- **Accessibility attributes** for products that have them: features, supported assistive technologies, compliance statements.
- **Accessible documents**: tagged PDFs or HTML versions of manuals and safety sheets.
- **Video assets** with captions and transcripts.
- **Readable text**: structured specifications in tables with headers, avoiding information embedded only in images.
- **Completeness rules** that flag missing alt text before publication, with named owners under [product data governance](https://pimatlas.com/knowledge/product-data-governance/).

Generative AI can draft alt text at scale, but review rules apply, as outlined in the guide to [generative AI for product descriptions](https://pimatlas.com/knowledge/generative-ai-product-descriptions/). Product images showing text, such as packaging, need descriptions that convey that text.

## Key takeaways

- Directive (EU) 2019/882 has applied since 28 June 2025 and covers e-commerce services offered to consumers.
- Online shops must meet the principles of perceivability, operability, understandability and robustness, which affects images, video and documents.
- Shops must show accessibility information about products when the responsible economic operator provides it.
- Microenterprises providing services are exempt; transitional provisions run to 28 June 2030 in defined cases.
- Treat alt text, accessible documents and accessibility attributes as governed PIM and DAM data.

Q: Does the European Accessibility Act apply to all online shops?
A: Directive (EU) 2019/882 applies to e-commerce services offered to consumers in the EU from 28 June 2025, as transposed into each Member State's law. Microenterprises providing services, meaning those with fewer than 10 employees and annual turnover or balance sheet not exceeding EUR 2 million, are exempt from the service accessibility requirements. National laws may differ in detail.
Q: What accessibility information must a webshop show about products?
A: Annex I, Section IV(g)(i) requires e-commerce services to provide the information concerning accessibility of the products and services being sold when this information is provided by the responsible economic operator. In practice, the shop must display accessibility information that manufacturers or suppliers supply, such as accessibility features or compatibility with assistive technologies.
Q: Is there a transition period under the European Accessibility Act?
A: Article 32 allows Member States to provide a transitional period ending on 28 June 2030 for services using products lawfully used before 28 June 2025. Service contracts agreed before that date may continue until they expire, for at most five years. Self-service terminals may remain in use until the end of their economically useful life, up to 20 years.

## Article: PIM for manufacturers: technical data, variants and dealers
URL: https://pimatlas.com/knowledge/pim-for-manufacturers/
Category: Industry playbooks | Updated: 2026-10-17

Summary: A PIM system gives manufacturers one governed source for technical attributes, variants, documents and translations, and publishes that data to dealers, distributors and webshops in the structures they require, such as ETIM-classified BMEcat files. The capabilities that decide fit are variant modelling, classification support, document management and partner-specific exports.

Manufacturers need a PIM (product information management) system when technical product data has to reach dealers, distributors, marketplaces and their own websites faster and more consistently than ERP exports and spreadsheets allow. The strongest case sits with firms that sell variant-heavy or configurable products through indirect channels, because every partner asks for the same data in a different structure.

## Why do manufacturers need a PIM system?

Engineering and ERP systems hold the data that defines how a product is built and sold: bills of materials, item numbers, prices and stock. They rarely hold what a buyer needs to select the product, such as classified technical attributes, application images, installation manuals and translations in every sales language. The handover from engineering to commerce is covered in [PIM and PLM](https://pimatlas.com/knowledge/pim-vs-plm/).

Three pressures make that gap more expensive in 2026:

- **Partner data requirements.** Wholesalers in electrical, HVAC, plumbing and industrial supply expect classified data in standard formats, and they rank or delist products with incomplete records.
- **Digital instructions.** The Machinery Regulation (EU) 2023/1230 applies from 20 January 2027. It allows instructions in digital form, with a paper copy supplied on request, so manuals become managed digital assets tied to models.
- **Sustainability data.** The Ecodesign for Sustainable Products Regulation (EU) 2024/1781 entered into force on 18 July 2024 and introduces the [Digital Product Passport](https://pimatlas.com/knowledge/digital-product-passport/), which product-group delegated acts will make mandatory step by step.

## How should manufacturers model technical data and variants?

A manufacturer's data model has to reflect how engineers define products and how buyers filter them. A pump range, for example, shares a family description and certificates, while each variant differs in flow rate, voltage, connection size and material. Modelling this as one family with defined variant axes avoids hundreds of near-duplicate records.

| Modelling decision | Recommended approach | Why it matters |
|---|---|---|
| Family level | Shared descriptions, images, certificates and documents | Edit once, inherit everywhere |
| Variant axes | Two or three axes that buyers actually select on | Keeps configurators and filters usable |
| Units and values | Numeric attributes with units, never free text | Required for ETIM and ECLASS exports |
| Documents | Typed assets (manual, datasheet, declaration of conformity) per language | Partners request documents by type |
| Lifecycle status | Explicit status for new, active, phase-out and spare part | Dealers need successor and spare-part links |

The detailed method for families, attributes and inheritance is set out in [product data modelling](https://pimatlas.com/knowledge/product-data-modelling/).

## Which classification standards matter for manufacturers?

Classification turns attributes into data that partner systems can read without manual mapping. Three standards dominate European B2B exchange, and they are explained further in [ETIM, BMEcat and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/).

- **ETIM** is a classification model for technical products that began in the Dutch electrotechnical sector and now covers HVAC, plumbing, building materials and shipbuilding. ETIM International coordinates it with member organisations representing 21 countries, and ETIM 10.0 was released on 5 December 2024.
- **ECLASS** describes itself as an ISO/IEC-compliant standard for classifying and describing products and services. Release 15.0, published in November 2024, added 490 classes and about 810 properties.
- **BMEcat** is an XML format for electronic catalogues that carries prices, documents and ETIM or ECLASS attribute values in one file.

A PIM should store one internal model and map it to each standard and version, because partners migrate between releases at different speeds.

## How do manufacturers serve dealers and distributors with product data?

Indirect sales put the manufacturer's data in front of buyers through someone else's webshop, so data quality directly affects sell-through. The practical requirements are:

1. Partner-specific export profiles (attribute sets, languages, file formats and delivery schedules per partner).
2. A dealer portal or self-service download for media, datasheets and price-independent product data.
3. Completeness rules per channel, so a record ships only when the partner's mandatory attributes are filled.
4. Change notifications, so dealers receive updates and discontinuations without a full reload.

When retailer and marketplace channels are added, the same principles extend into [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/).

## What should manufacturers test during PIM selection?

Scripted demos with your own data reveal fit faster than feature lists. Ask each vendor to load one variant-heavy family, export it as an ETIM-classified BMEcat file, attach multilingual manuals, and show how an engineering change from ERP or PLM flows through to a dealer export. Shortlisted products can be checked side by side in the [compare tool](https://pimatlas.com/compare/).

## Key takeaways

- Manufacturers gain most from PIM when they sell variant-heavy products through dealers and distributors.
- Model families with two or three real variant axes, numeric attributes with units and typed documents.
- Maintain ETIM and ECLASS mappings from one internal model and export them through BMEcat.
- Treat digital instructions under Regulation (EU) 2023/1230, applicable from 20 January 2027, as versioned product assets.
- Test vendors with your own variant data and a real partner export.

Q: Do manufacturers need a PIM if they already have an ERP and a PLM system?
A: Usually yes, once they sell through several channels. ERP holds commercial master data such as item numbers, prices and stock, and PLM holds engineering definitions. Neither is designed to manage marketing copy, classified sales attributes, media, translations and channel-specific exports. A PIM takes approved data from both systems, enriches it and publishes it to dealers, distributors and websites.
Q: Which classification standard should a manufacturer use, ETIM or ECLASS?
A: Follow the channel. ETIM dominates electrical, HVAC, plumbing and building-material wholesale in Europe, coordinated by ETIM International with member organisations in 21 countries. ECLASS is stronger in industrial procurement and engineering, and describes itself as an ISO/IEC-compliant standard. Many manufacturers maintain both mappings in their PIM and export each one to the partners that request it.
Q: How does the EU Machinery Regulation affect product information for manufacturers?
A: Regulation (EU) 2023/1230 applies from 20 January 2027 and allows manufacturers to supply instructions in digital form, with a paper copy available when the customer requests it. That turns manuals into versioned digital assets linked to specific machine models, which a PIM or connected DAM can manage, publish and keep retrievable for the required period.

## Article: PIM for wholesale and distribution: supplier onboarding at scale
URL: https://pimatlas.com/knowledge/pim-for-wholesale-distribution/
Category: Industry playbooks | Updated: 2026-10-17

Summary: Wholesalers and distributors use PIM to ingest product data from hundreds of suppliers, normalise it to one classification model such as ETIM, score its quality and publish catalogues of hundreds of thousands of items. The deciding capabilities are automated supplier onboarding, mapping to standards, rule-based quality gates and performance at large SKU volumes.

Wholesalers and distributors need a PIM (product information management) system built for inbound data: the core job is collecting product information from many suppliers, normalising it into one model and publishing it to webshops, print catalogues and customer systems. Selection should focus on supplier onboarding, standards mapping, quality gates and performance at very large catalogue sizes.

## Why is PIM different for wholesale and distribution?

A manufacturer creates its own product data. A distributor receives it, often from several hundred suppliers, each with its own attribute names, units, languages and file formats. The distributor adds value by making that data consistent, searchable and comparable across brands.

That shifts the requirements:

- **Volume.** Catalogues of several hundred thousand to millions of items are normal, so import speed, bulk editing and search performance matter more than editorial features.
- **Heterogeneity.** The same attribute arrives as "Spannung", "voltage", "V" or a value inside a description field.
- **Shared ownership.** Suppliers own the source facts, while the distributor owns classification, descriptions, assortment and channel decisions.
- **Customer-facing outputs.** Professional buyers consume the result in webshops, printed and PDF catalogues, procurement systems and customer-specific data exports, each with its own field list.

## How should distributors onboard supplier data at scale?

Effective onboarding uses several inbound channels and one controlled landing zone. The table summarises the usual options.

| Inbound channel | Best for | Typical standard | Distributor effort |
|---|---|---|---|
| Standard catalogue file | Technical suppliers in electrical, HVAC, tools | BMEcat with ETIM or ECLASS | Low once mapped |
| Data pool synchronisation | Consumer goods, food, healthcare | GS1 GDSN | Low for GS1 attributes |
| Supplier portal with templates | Small and mid-sized suppliers | Distributor's own template | Medium |
| AI-assisted extraction | Long tail of PDFs and spreadsheets | None | Medium, needs review |

GDSN (Global Data Synchronisation Network) connects trading partners through GS1-certified data pools linked to the GS1 Global Registry, so updates from a supplier flow to subscribers automatically. BMEcat is an XML catalogue format that carries prices, documents and ETIM or ECLASS attribute values. Both are covered in [GS1, GTINs and GDSN](https://pimatlas.com/knowledge/gs1-gtin-gdsn/) and [ETIM, BMEcat and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/).

Whatever the channel, incoming data should land in a staging area, pass mapping and validation rules, and only then update the live catalogue.

## How do distributors normalise and classify supplier data?

Normalisation turns supplier data into the distributor's model. A practical sequence:

1. Match the item to an existing record using GTIN or manufacturer part number, to prevent duplicates across suppliers.
2. Classify it into the distributor's taxonomy, typically aligned to an ETIM class in technical sectors.
3. Map supplier attributes and units to the target features, keeping the mapping per supplier so the next delivery runs automatically.
4. Apply value lists, so "stainless steel", "RVS" and "inox" become one value.
5. Flag gaps for supplier follow-up or AI-assisted enrichment.

ETIM, coordinated by ETIM International with member organisations in 21 countries, gives distributors in technical sectors a shared target model. ETIM 10.0 was released on 5 December 2024, so distributors should state which version they accept.

## How do distributors measure and enforce data quality?

Quality has to be measured per channel and per supplier, because a record can be complete for a print catalogue and incomplete for a filterable webshop. Useful metrics include completeness of mandatory features per class, image coverage, document coverage and time from supplier delivery to publication. Methods for scoring are described in [product data quality](https://pimatlas.com/knowledge/product-data-quality/).

Regulation adds mandatory fields. Under the General Product Safety Regulation (EU) 2023/988, applicable from 13 December 2024, online offers must show manufacturer identity and contact details, an EU responsible person where relevant, product identification and safety warnings. Distributors should make these attributes mandatory in onboarding and block publication when they are missing.

## What should distributors test in a PIM selection?

Ask vendors to import a real BMEcat file and a messy supplier spreadsheet, map both to one ETIM class, show duplicate detection, and run a bulk update across 100,000 records. Then check how the result flows into the webshop, which links to [PIM for B2B e-commerce](https://pimatlas.com/knowledge/pim-for-b2b-ecommerce/). Finally, ask for reference customers with a comparable supplier count and catalogue size.

## Key takeaways

- Distributor PIM is an inbound problem: collect, normalise, score, then publish.
- Combine BMEcat, GDSN, a supplier portal and AI-assisted extraction, all landing in one staging area.
- Keep per-supplier mappings to ETIM or another target model so repeat deliveries run automatically.
- Make GPSR fields under Regulation (EU) 2023/988 mandatory before publication.
- Test performance and bulk operations at your real catalogue size.

Q: How do distributors onboard product data from hundreds of suppliers?
A: They combine three channels: standard feeds where suppliers can deliver them (BMEcat files with ETIM or ECLASS attributes, GDSN for GS1-based categories), a supplier portal with templates and validation for the rest, and AI-assisted extraction from PDFs and spreadsheets for the long tail. Every channel lands in a staging area where mapping and quality rules run before data reaches the live catalogue.
Q: Should a distributor require suppliers to deliver ETIM data?
A: In electrical, HVAC, plumbing, tools and building materials, requiring ETIM-classified data is common practice and cuts mapping effort sharply. ETIM International coordinates the model with member organisations in 21 countries. Distributors should state the accepted ETIM version, the mandatory features per class and the delivery format, usually BMEcat, in their supplier data requirements.
Q: Who is responsible for GPSR product information when a distributor sells online?
A: Under Regulation (EU) 2023/988, applicable from 13 December 2024, an offer made through distance sales must clearly show the manufacturer's identity and contact details, an EU responsible person where the manufacturer is outside the EU, product identification, and warnings or safety information. A distributor publishing those offers needs that data from suppliers and should treat it as mandatory in onboarding.

## Article: PIM for retail and fashion: speed, seasons and marketplaces
URL: https://pimatlas.com/knowledge/pim-for-retail-fashion/
Category: Industry playbooks | Updated: 2026-10-17

Summary: Retailers and fashion brands use PIM to launch large seasonal collections on time, manage style, colour and size variants, and feed webshops and marketplaces with channel-ready content. The deciding capabilities are variant modelling, image-led workflows with DAM, marketplace syndication and readiness for EU fibre composition and Digital Product Passport data.

Retailers and fashion brands need a PIM (product information management) system to get large, short-lived assortments online on time and in a consistent form across their own webshop, apps and marketplaces. The selection criteria that matter most are variant modelling, image-led workflows, marketplace syndication and readiness for EU textile data obligations.

## Why do retail and fashion need a different PIM approach?

Fashion product data has a short shelf life. A brand may launch thousands of new styles per season, each sold for a few months, and revenue lost to a late or incomplete listing does not come back. Speed to publication is the primary metric.

Three traits shape the requirements:

- **Variant depth.** One style can expand into dozens of colour and size combinations, each with its own GTIN.
- **Image dependence.** Shoppers choose on visuals, so photography and retouching usually sit on the critical path.
- **Channel spread.** Marketplaces and retail partners each impose their own taxonomy, size charts and image rules.

## How should fashion product data be modelled?

The standard model has three levels, which mirrors how buyers browse and how marketplaces expect parent-child listings. The general method is covered in [product data modelling](https://pimatlas.com/knowledge/product-data-modelling/).

| Level | Typical attributes | Owner |
|---|---|---|
| Style | Name, description, fibre composition, care, fit, season, category | Merchandising and copy |
| Colour | Colour name and family, images, video, colour-specific copy | Studio and e-commerce |
| Size | GTIN, size code, size system, measurements, stock link | Product data and ERP |

Size systems deserve explicit treatment. Storing size, size system (EU, UK, US, alpha) and measurements as separate attributes lets the PIM generate the size labels and conversion tables each channel requires.

## How can PIM shorten time to market for seasonal collections?

The goal is to start enrichment before samples arrive and finish it the moment images are approved. A practical seasonal workflow:

1. Import the line plan from PLM or a buying tool, creating styles with basic attributes weeks before launch.
2. Generate first-draft copy and attribute suggestions with AI, routed to a copywriter for review.
3. Link images from the DAM automatically by naming convention or SKU, as described in [integrating DAM and PIM](https://pimatlas.com/knowledge/dam-pim-integration/).
4. Gate publication on channel completeness: images, composition, care and size data present.
5. Publish per channel and per market on scheduled launch dates.

Teams that track hours from image approval to live listing can see where this chain stalls. Image specifications per channel are covered in [product image standards](https://pimatlas.com/knowledge/product-image-standards/).

## How do retailers feed marketplaces and partners from PIM?

Each marketplace has its own category tree, mandatory attributes and value lists. Maintaining those mappings inside the PIM, or in a connected syndication tool, keeps the master record clean while each channel receives what it requires. Mapping and feed strategies are compared in [product data syndication](https://pimatlas.com/knowledge/product-data-syndication/).

Two developments add requirements in 2026 and 2027:

- **Accessibility.** Directive (EU) 2019/882, the European Accessibility Act, applies to e-commerce services provided to consumers after 28 June 2025. Alt text for product images and accessible size and composition information are now compliance topics.
- **2D barcodes.** GS1 US leads Sunrise 2027, an initiative to prepare retail point-of-sale systems to scan 2D barcodes by the end of 2027. Brands using QR codes based on GS1 Digital Link need GTIN-level data and web links that resolve correctly.

## Which EU rules shape fashion product data?

Regulation (EU) No 1007/2011 requires textile fibre composition to be clearly visible before purchase, including for purchases made by electronic means. That makes a structured composition attribute, using the legally defined fibre names, a requirement for every online listing.

The Ecodesign for Sustainable Products Regulation (EU) 2024/1781 entered into force on 18 July 2024 and bans the destruction of unsold apparel and footwear. Textile apparel is a priority group for a [Digital Product Passport](https://pimatlas.com/knowledge/digital-product-passport/). The European Commission indicates adoption of the textile delegated act in Q4 2027 and expects the passport to cover fibre composition, origin, care and repair information and economic operator identification. Much of that data already sits in a fashion PIM, which is the logical place to structure it.

## Key takeaways

- Fashion PIM success is measured in time from image approval to live listing.
- Model style, colour and size as separate levels, with images at colour level and GTINs at size level.
- Keep marketplace mappings outside the master record, in the PIM or a syndication layer.
- Treat fibre composition under Regulation (EU) No 1007/2011 as mandatory structured data.
- Prepare composition, origin and care data now for the textile Digital Product Passport.

Q: How should a fashion PIM model style, colour and size?
A: Most fashion teams use a three-level model: the style holds shared copy, composition and care; the colour level holds images and colour names; the size level holds GTINs, size codes and measurements. Images belong at colour level because shoppers choose colours visually. This structure prevents duplicate copy and lets marketplaces receive the parent-child relationships they require.
Q: What fibre composition information must online fashion retailers show in the EU?
A: Regulation (EU) No 1007/2011 on textile fibre names and labelling requires the fibre composition to be clearly visible to the consumer before purchase, including when the purchase is made by electronic means. In practice, every product detail page and marketplace listing needs a structured, correctly named fibre composition attribute managed in the PIM.
Q: When will the Digital Product Passport apply to clothing?
A: Textile apparel is a priority product group in the ESPR Working Plan 2025-2030. The European Commission's indicative timeline shows adoption of the delegated act with ecodesign and passport requirements in Q4 2027, with application after a transition period. The Commission notes the timeline may shift, so brands should prepare structured composition, origin and care data now.

## Article: PIM for B2B e-commerce: catalogues, pricing context and punchout
URL: https://pimatlas.com/knowledge/pim-for-b2b-ecommerce/
Category: Industry playbooks | Updated: 2026-10-17

Summary: In B2B e-commerce, PIM supplies the searchable, classified product content, while ERP or a pricing engine supplies customer-specific prices and stock. PIM also defines customer assortments and feeds punchout catalogues through SAP OCI or cXML. The deciding capabilities are filterable technical attributes, assortment logic, clean ERP integration and procurement-ready exports.

B2B e-commerce needs a PIM (product information management) system to deliver the specification-level product content professional buyers search on, organised into customer-specific assortments and exported to procurement systems. Prices and stock stay in ERP or a pricing engine. The deciding capabilities are filterable technical attributes, assortment logic, clean ERP integration and punchout support.

## How does B2B e-commerce change PIM requirements?

Business buyers rarely browse. They arrive with a part number, a specification or a reorder list, and they buy on contract terms. That shifts the job of product content from persuasion to precise identification.

Four differences from B2C stand out:

- **Specification search.** Faceted filters on voltage, diameter, material or pressure rating only work when attributes are numeric, unit-normalised and consistently classified.
- **Account-specific views.** Each customer may see a different assortment, product names and part numbers of their own, and contract prices.
- **Order units.** Packaging quantity, minimum order quantity and price unit matter as much as the product itself.
- **System-to-system buying.** Many large buyers order from inside SAP, Coupa or other procurement suites rather than on the webshop.

## Which product data belongs in PIM and which in ERP?

Clear ownership prevents the most common B2B integration failures. The full reasoning is set out in [PIM and ERP](https://pimatlas.com/knowledge/pim-vs-erp/).

| Data | System of record | Delivered to the shop |
|---|---|---|
| Item number, order units, packaging quantity | ERP | Via PIM or directly |
| Technical attributes, classification, descriptions | PIM | From PIM |
| Images, datasheets, certificates | DAM or PIM | From PIM or DAM |
| Customer assortments and customer part numbers | PIM or commerce platform | From PIM or commerce platform |
| Contract prices, tiered prices, discounts | ERP or pricing engine | Real-time API call |
| Stock and delivery dates | ERP or warehouse system | Real-time API call |

The pattern is simple: slow-changing content flows from PIM in batches or events, while fast-changing commercial data is called live. Integration patterns are compared in [PIM integration architecture](https://pimatlas.com/knowledge/pim-integration-architecture/).

## How does PIM support punchout catalogues?

A punchout catalogue lets a buyer leave a procurement system, shop in the supplier's catalogue and return the basket as a requisition for approval. Two protocols dominate:

1. **OCI (Open Catalog Interface)**, developed by SAP, returns basket lines as HTTP form fields to a return address the buyer's system supplies. It is used mainly with SAP Supplier Relationship Management.
2. **cXML (commerce XML)**, created by Ariba, uses XML messages for punchout sessions and also covers purchase orders, order confirmations, ship notices and invoices.

The PIM's role is to supply clean descriptions, classification codes, units of measure and images to the punchout catalogue, and to produce static catalogue files for buyers who load content into their own procurement systems. Buyers often request classification codes such as UNSPSC or ECLASS on every line, so these mappings should live in the PIM.

## What product data do B2B buyers search on?

Specification data drives findability. Distributors and manufacturers in technical sectors typically align attributes with ETIM, coordinated by ETIM International with member organisations representing 21 countries, or with ECLASS. The standards are compared in [ETIM, BMEcat and ECLASS](https://pimatlas.com/knowledge/etim-bmecat-eclass/).

A practical minimum per item:

- Manufacturer name and manufacturer part number
- GTIN where one exists
- Classified technical attributes with units
- Order unit, packaging quantity and minimum order quantity
- Datasheet, certificates and safety documents
- Accessory, spare part and successor relationships

The same structured data also feeds search engines and AI assistants. The schema.org Product type supports properties such as gtin, mpn, sku and brand, and AI shopping agents rely on the same fields, as discussed in [agentic commerce](https://pimatlas.com/knowledge/agentic-commerce-product-data/).

## What should B2B companies test during PIM selection?

Ask vendors to show one customer-specific assortment, a customer part number mapping, a live price call from ERP alongside PIM content, and a punchout or catalogue export with classification codes. Then measure how long a new SKU takes from ERP creation to searchable listing. Shortlisted vendors can be checked side by side in the [compare tool](https://pimatlas.com/compare/), and a lead time measured in days rather than hours usually points to manual handoffs in the integration.

## Key takeaways

- B2B PIM supplies specification content and assortments, while ERP or a pricing engine supplies prices and stock.
- Numeric, unit-normalised, classified attributes are the foundation of B2B search.
- Support both SAP OCI and Ariba cXML punchout, with classification codes on every line.
- Keep customer part numbers and assortments in a governed system, usually the PIM or commerce platform.
- Test with a real customer scenario that combines PIM content and live ERP pricing.

Q: Should prices be managed in the PIM for B2B e-commerce?
A: Usually not. Customer-specific prices, contract discounts and tiered pricing change often and depend on the account, so they belong in ERP or a dedicated pricing engine. The PIM can hold price-related context such as list price, price unit, packaging quantity and minimum order quantity, which the commerce platform combines with live prices at request time.
Q: What is the difference between OCI and cXML punchout?
A: Both let a buyer leave a procurement system, shop in a supplier's catalogue and return the basket as a requisition. OCI (Open Catalog Interface) was developed by SAP and returns basket items as HTTP form fields. cXML (commerce XML) was created by Ariba and uses XML messages, covering punchout sessions, purchase orders, order confirmations and invoices. Many suppliers support both.
Q: What product data do B2B buyers need to find the right item?
A: Professional buyers search by specification. They need numeric technical attributes with units, classification aligned with ETIM or ECLASS, manufacturer part numbers, packaging and order units, datasheets and certificates, and accessory and successor links. Without these, faceted search fails and buyers fall back to phone and email orders.

## Glossary
- 3D asset: A 3D asset is a digital three-dimensional model of a product, including geometry, materials and textures, used for interactive viewers, augmented reality, configurators and rendered images. Product 3D assets are usually created from CAD data or scans and managed in a DAM with renditions in delivery formats such as glTF and USDZ.
- Agentic commerce: Agentic commerce is buying and selling in which AI agents act on behalf of users, for example by searching for products, comparing offers, checking specifications and completing purchases with limited human input. It places high demands on product data: machine-readable attributes, accurate prices and availability, clear identifiers and structured policies.
- AI Act: The AI Act is Regulation (EU) 2024/1689, the European Union's horizontal law on artificial intelligence, which classifies AI systems by risk and sets obligations for providers and deployers. The Digital Omnibus on AI, Regulation (EU) 2026/1744, postponed obligations for stand-alone high-risk systems to 2 December 2027 and for high-risk systems embedded in products to 2 August 2028.
- API-first: API-first is a design approach in which a software product exposes all of its functions through documented application programming interfaces before, or alongside, any user interface. For PIM and DAM, an API-first design means that integrations, automation and custom front ends can read and write the same data and actions available to users.
- Attribute: An attribute is a single, named characteristic of a product, such as width, voltage, colour, material or marketing description. In a PIM, each attribute has a data type, optional unit, validation rules and sometimes a value list. Well-defined attributes make product data filterable, comparable and exportable to classification standards and channel templates.
- Auto-tagging: Auto-tagging is the automatic assignment of keywords, categories or attributes to digital assets using computer vision and machine learning, for example recognising product type, colour, scene or text in an image. In DAM systems, auto-tagging speeds up metadata entry and search, but tags still require vocabulary control and spot checks for accuracy.
- BMEcat: BMEcat is an XML-based standard for exchanging electronic product catalogues between suppliers and buyers, widely used in German-speaking and European B2B markets. A BMEcat file carries product identifiers, descriptions, prices, media references and classified attributes, often using ETIM or ECLASS. The most common version in use is BMEcat 2005.
- C2PA (Coalition for Content Provenance and Authenticity): C2PA (Coalition for Content Provenance and Authenticity) publishes a technical specification for recording where digital content came from and how it was changed. Signed statements about an asset are bound to it with cryptographic hashes in a C2PA manifest, stored as Content Credentials. DAM teams use it to label AI-generated or edited product imagery.
- CDN (Content delivery network): A CDN (content delivery network) is a distributed network of servers that caches and delivers files, such as product images and videos, from locations close to the end user. DAM and PIM platforms use CDNs to serve renditions quickly to webshops and apps worldwide, often combined with on-the-fly image resizing and format conversion.
- Channel: A channel is any destination through which product information reaches buyers, such as an own webshop, mobile app, marketplace, retailer, distributor, print catalogue, point of sale or procurement system. In a PIM, each channel can have its own attribute set, locales, completeness rules, asset formats and export schedule.
- Completeness score: A completeness score is a metric showing what share of the required attributes, media and documents for a product is filled for a given channel and locale, usually expressed as a percentage. PIM systems calculate it from rules per family and channel, and teams use it to gate publication and to prioritise enrichment work.
- Composable commerce: Composable commerce is an approach to building digital commerce from independent, interchangeable components, such as PIM, search, cart, checkout, content management and pricing, selected from different vendors and connected through APIs. It gives flexibility to change individual parts, but requires stronger integration, orchestration and governance capabilities than a single suite.
- Content syndication: Content syndication is the distribution of rich product content, such as enhanced descriptions, image galleries, videos, comparison charts and feature tiles, to retailer and marketplace product pages. It complements syndication of core attributes and is often delivered through retailer-specific templates or embedded content services rather than standard data feeds.
- cXML (Commerce eXtensible Markup Language): cXML (commerce XML) is an XML-based protocol created by Ariba for exchanging business documents between procurement applications, e-commerce hubs and suppliers. It covers punchout sessions, purchase orders and changes, order confirmations, ship notices, invoices and catalogue content, and is widely supported by procurement suites and supplier e-commerce platforms.
- DAM (Digital asset management): DAM (digital asset management) is software for storing, describing, finding, approving and distributing digital files such as product photos, videos, 3D models, documents and brand assets. A DAM system manages metadata, versions, usage rights and renditions, and typically supplies approved product media to a PIM, website or marketing tools through links or APIs.
- Data pool: A data pool is a service that stores and exchanges product data on behalf of suppliers and recipients within GS1's Global Data Synchronisation Network. Each data pool must be certified and tested by GS1 and connects to the GS1 Global Registry. A company needs to join only one certified data pool to reach trading partners on any other pool.
- Data quality rule: A data quality rule is a defined, testable condition that product data must meet, such as a mandatory attribute, an allowed value range, a valid GTIN check digit or a minimum image resolution. PIM systems run these rules automatically on import or edit, report violations, and can block publication of records that fail them.
- Data steward: A data steward is the person accountable for the quality, definitions and correct use of a defined set of data, such as the product attributes for one category or region. In product data governance, data stewards maintain attribute definitions and value lists, resolve data issues, approve changes and monitor quality metrics.
- Digital Product Passport (DPP): A Digital Product Passport (DPP) is a structured digital record of a product's sustainability, circularity and compliance information, established by the Ecodesign for Sustainable Products Regulation (EU) 2024/1781. It is linked to a unique product identifier and accessed through a data carrier such as a QR code. Delegated acts define its content per product group.
- EAA (European Accessibility Act): The EAA (European Accessibility Act) is Directive (EU) 2019/882 on accessibility requirements for products and services. Member States apply its measures from 28 June 2025, and it covers e-commerce services provided to consumers. For product content, it raises requirements for text alternatives for images, accessible product information and accessible online shops.
- ECLASS: ECLASS is a classification and product description standard for products and services, maintained by ECLASS e.V. and described by its publisher as ISO/IEC-compliant. It defines classes, properties, values and units in a machine-readable form, and is widely used in industrial procurement, engineering and Industry 4.0 applications. Release 16.0 is the current version in 2026.
- Enrichment: Enrichment is the process of adding the information needed to sell a product to its basic record, such as descriptions, technical attributes, classifications, images, documents, translations and relationships. Enrichment may be manual, rule-based or AI-assisted, and is usually organised as a workflow with defined owners, quality checks and approval before publication.
- ERP (Enterprise resource planning): ERP (enterprise resource planning) is the core transactional system for finance, purchasing, inventory, production and order management. For product data, ERP is usually the system of record for item numbers, units of measure, costs, prices and stock. It rarely holds the rich descriptive content, media and classified attributes that a PIM manages for sales channels.
- ESPR (Ecodesign for Sustainable Products Regulation): ESPR (Ecodesign for Sustainable Products Regulation) is Regulation (EU) 2024/1781, which entered into force on 18 July 2024 and repealed the Ecodesign Directive 2009/125/EC. It sets a framework for product-group ecodesign requirements, introduces the Digital Product Passport and bans the destruction of unsold apparel and footwear. Delegated acts apply it to specific product groups.
- ETIM (ElectroTechnical Information Model): ETIM is an open classification standard for technical products that defines product classes with their features, values and units. It began in the Dutch electrotechnical sector and now covers HVAC, plumbing, building materials, tools and shipbuilding. ETIM International coordinates it with member organisations representing 21 countries, and ETIM 10.0 was released on 5 December 2024.
- GDSN (Global Data Synchronisation Network): GDSN (Global Data Synchronisation Network) is a GS1 network that lets trading partners share standardised product data automatically. Suppliers publish data through a GS1-certified data pool, which connects to the GS1 Global Registry, and subscribing retailers receive updates through their own data pool. It is widely used in grocery, consumer goods and healthcare.
- Generative AI: Generative AI is a category of artificial intelligence that creates new content, such as text, images, audio or code, based on patterns learned from training data. In product content, generative AI drafts descriptions, titles, translations and image variations, which teams then review against brand, accuracy and regulatory requirements before publication.
- GEO (Generative engine optimisation): GEO (generative engine optimisation) is the practice of structuring and writing content so that AI assistants and AI-driven search engines can find, understand, trust and cite it in generated answers. For product content, GEO relies on complete structured attributes, clear definitions, schema.org markup, consistent identifiers and factual, quotable descriptions.
- glTF (GL Transmission Format): glTF (GL Transmission Format) is a Khronos Group format for efficient transmission and loading of 3D content in applications and on the web. Version glTF 2.0 was published as the international standard ISO/IEC 12113:2022 in July 2022. It is widely used for web-based 3D product viewers and augmented reality.
- Golden record: A golden record is the single, consolidated and trusted version of a data entity, such as a product, created by matching, merging and validating data from several source systems. In MDM, survivorship rules decide which source wins for each attribute. Downstream systems then consume the golden record instead of maintaining their own conflicting copies.
- GPSR (General Product Safety Regulation): GPSR (General Product Safety Regulation) is Regulation (EU) 2023/988, which applies from 13 December 2024. For distance sales, Article 19 requires offers to show the manufacturer's identity and contact details, an EU responsible person where relevant, information identifying the product, and any warnings or safety information. These become mandatory fields in PIM.
- GS1 Digital Link: GS1 Digital Link is a GS1 standard for expressing GS1 identifiers, such as the GTIN, batch or serial number, as web addresses. Encoded in a QR code or other carrier, one GS1 Digital Link can serve point-of-sale scanning and link consumers to online information. QR codes carrying it are one of the 2D options retailers are preparing to scan under Sunrise 2027.
- GTIN (Global Trade Item Number): A GTIN (Global Trade Item Number) is a GS1 identifier for trade items, used to identify products uniquely across companies and channels. GTINs have 8, 12, 13 or 14 digits and are encoded in barcodes such as EAN-13 and UPC-A. GS1 allocates a company prefix to each member, which then assigns item references to its products.
- Headless: Headless describes an architecture in which the back-end system that stores and manages content or commerce logic is separated from the front-end presentation layer, and the two communicate through APIs. A headless PIM or commerce platform supplies product data to any number of front ends, such as websites, apps, kiosks and marketplaces.
- IPTC (International Press Telecommunications Council): The IPTC Photo Metadata Standard, published by the International Press Telecommunications Council, defines descriptive, administrative and rights properties for images, such as creator, caption, keywords, copyright notice and licensing details. Its values are embedded in image files using XMP and the older IIM format. Version 2025.1 is current.
- LLM (Large language model): An LLM (large language model) is a neural network trained on very large text collections to predict and generate language. LLMs power chat assistants and many PIM features, such as description writing, attribute extraction, classification suggestions and translation. Their output can contain errors, so product teams pair them with source data, rules and human review.
- Localisation: Localisation is the adaptation of product content to a specific market, covering translation as well as units of measure, size systems, currencies, legal statements, cultural references and channel conventions. A PIM supports localisation through locale-specific attributes, inheritance from a base language, translation workflows and completeness checks per locale.
- MACH (Microservices, API-first, Cloud-native, Headless): MACH is an architecture principle combining microservices, API-first design, cloud-native deployment and headless delivery, promoted by the not-for-profit MACH Alliance founded in 2020. In product data, MACH-aligned PIM, DAM and commerce services are deployed independently and connected through APIs, so each component can be replaced without rebuilding the whole stack.
- Marketplace: A marketplace is an online platform on which third-party sellers list and sell products to buyers, with the platform operating the storefront, search and often payment. Marketplaces impose their own category trees, mandatory attributes, value lists and image rules, so sellers usually map PIM data to each marketplace's requirements through feeds or APIs.
- MDM (Master data management): MDM (master data management) is the discipline of creating and maintaining a single, trusted version of core business entities such as products, customers, suppliers and locations across enterprise systems. Product MDM focuses on identifiers, hierarchies and operational attributes shared by ERP, supply chain and finance, while PIM focuses on the commercial content used to sell products.
- Metadata: Metadata is structured information that describes other data or files, such as the title, keywords, product codes, photographer, rights and expiry date attached to an image. In DAM and PIM, consistent metadata makes assets searchable, links them to the right products and controls where and how long they may be used.
- OCI (Open Catalog Interface): OCI (Open Catalog Interface) is a punchout protocol developed by SAP that connects a buyer's procurement system to a supplier's online catalogue. After the buyer checks out, the catalogue posts the basket lines, including description, supplier part number, quantity, unit and price, as HTTP form fields to a return address supplied by the buyer's system.
- PIM (Product information management): PIM (product information management) is the discipline and software used to collect, enrich, govern and publish the marketing and technical information that describes products. A PIM system holds attributes, descriptions, classifications, relationships and links to media in one governed repository, and distributes channel-ready product data to webshops, marketplaces, print, retailers and distributors.
- PLM (Product lifecycle management): PLM (product lifecycle management) is software that manages a product from concept and engineering through manufacturing, change control and retirement. PLM holds specifications, bills of materials, CAD references and engineering changes. In a typical architecture, approved PLM data is handed to ERP for operations and to PIM for commercial enrichment and publication.
- Product content supply chain: The product content supply chain is the end-to-end flow of product information and media from creation to consumption, covering engineering or supplier input, enrichment, asset production, approval, localisation, syndication and performance feedback. Managing it as a supply chain means measuring lead times, quality and bottlenecks at every step.
- Product family: A product family is a group of products that share the same set of attributes and often the same descriptive content, such as all office chairs or all LED drivers. In a PIM, the family defines which attributes apply and which are mandatory, and serves as the parent from which variants inherit shared values.
- Product feed: A product feed is a structured file or data stream containing product records formatted to the specification of a receiving channel, such as a shopping engine, marketplace or advertising platform. Feeds usually include identifiers, titles, descriptions, prices, availability, images and category mappings, and are refreshed on a schedule or by API.
- Punchout: Punchout is a B2B purchasing method in which a buyer leaves their procurement system, shops in a supplier's online catalogue and returns the selected basket to the procurement system as a requisition for approval. The main protocols are SAP's Open Catalog Interface (OCI) and cXML. Punchout keeps catalogue content and prices under the supplier's control.
- PXM (Product experience management): PXM (product experience management) is an approach to product content that extends PIM with channel-specific storytelling, context and optimisation. It covers tailoring descriptions, media and attributes to each channel, audience and market, and measuring how that content performs. Several vendors use the term for PIM products that add syndication, content scoring and analytics.
- RAG (Retrieval-augmented generation): RAG (retrieval-augmented generation) is a technique in which an AI system retrieves relevant documents or records from a defined source, such as a PIM or knowledge base, and supplies them to a language model as context for its answer. RAG grounds generated text in approved product data and reduces invented facts.
- Rendition: A rendition is a derived version of a digital asset produced for a specific use, such as a thumbnail, a web-optimised JPEG, a WebP file, a marketplace image at a required size or a print-resolution TIFF. DAM systems generate renditions automatically from a master file, so channels receive correctly formatted files without manual export.
- Rights management: Rights management is the control of how digital assets may be used under licences, contracts and consents, covering permitted channels, territories, time periods, credits and model or property releases. In a DAM, rights metadata and expiry dates restrict downloads and distribution, and trigger alerts before an image or video may no longer be used.
- Schema.org Product: Schema.org Product is a structured data type from the schema.org vocabulary for describing any offered product or service on a web page. Properties include name, brand, sku, gtin, mpn, offers and aggregateRating, typically published as JSON-LD. Search engines and AI systems read this markup to understand product pages and show product details.
- SKU (Stock keeping unit): A SKU (stock keeping unit) is an identifier a company assigns internally to each distinct item it stocks or sells, such as one size and colour of a garment. SKUs are merchant-specific, so the same product carries different SKUs at different companies. Global identification across trading partners uses the GTIN, while the SKU supports internal inventory and ordering.
- Supplier onboarding: Supplier onboarding is the process of collecting, validating and importing product data and assets from suppliers into a retailer's or distributor's systems. It typically combines standard feeds such as GDSN or BMEcat, supplier portals with templates and validation, and mapping rules that convert each supplier's structure into the receiving company's data model.
- Syndication: Syndication is the automated distribution of product data and assets from a company's systems to external channels such as retailers, marketplaces, distributors and data pools. Syndication tools map the master data to each recipient's taxonomy, attributes and formats, validate it against their rules and deliver it by API, file or network.
- Taxonomy: A taxonomy is a hierarchical scheme that organises products into categories and subcategories, such as a navigation tree on a webshop or an internal classification. In product data, companies often maintain several taxonomies in parallel: a master classification that drives attributes, navigation trees per channel, and mappings to external standards such as ETIM, ECLASS or UNSPSC.
- Translation memory: A translation memory is a database of previously translated text segments paired with their source text, used by translation tools to reuse approved translations for repeated or similar content. In product data, translation memories reduce cost and keep terminology consistent across thousands of similar descriptions, attribute values and documents.
- UNSPSC (United Nations Standard Products and Services Code): UNSPSC (United Nations Standard Products and Services Code) is an open, global, multi-sector standard for classifying products and services in a hierarchical code set. GS1 US manages it on behalf of the United Nations Development Programme. It is used mainly in procurement and spend analysis, where buyers ask suppliers for a UNSPSC code per catalogue line.
- Variant: A variant is a version of a product that differs from its siblings on one or more defined axes, such as size, colour, voltage or capacity. Variants inherit shared content from a parent product or model and carry their own identifiers, such as a GTIN, and axis-specific values. Good variant modelling avoids duplicate records and supports channel parent-child listings.
- Webhook: A webhook is an automated HTTP message that one system sends to another when a defined event occurs, such as a product being approved or an asset being updated. Webhooks let PIM and DAM systems notify downstream systems in near real time, which reduces the need for frequent polling or large scheduled exports.
- Workflow: A workflow is a defined sequence of tasks, roles and approval steps that a product record passes through, such as creation, enrichment, translation, legal review and publication. PIM workflows assign tasks to users or groups, track status and deadlines, and ensure that only approved data reaches sales channels.
- XMP (Extensible Metadata Platform): XMP (Extensible Metadata Platform) is a framework for embedding metadata in digital files, such as images, PDFs and videos, so that the information travels with the file. Adobe introduced it in 2001, and its core data model was standardised as ISO 16684-1 in 2012. IPTC photo metadata is commonly stored using XMP.