PIMvendors.com B.V. publishes the PIM Report 2027. This edition is dated December 2026. Research closed on 17 October 2026, except where a page states a later access date.

PIMvendors.com is an independent PIM advisory. We do not rank vendors. Inclusion in this report is not an endorsement.

Vendor statements are reported as vendor claims. Summaries of regulation are not legal advice. Market sizes are estimates published by other firms, and they do not agree with each other. Where we draw a conclusion of our own, the page says so.

Foundations

1. 01 What a PIM is and doesThe product record, the journey it takes, and the line between what every PIM must do and what some add.
2. 02 When you need a PIMFive ways a spreadsheet runs out, and what poor product data actually costs.

Market and technology

1. 03 The PIM marketEvery publisher agrees the market grows. They disagree on almost everything else, and the reasons matter.
2. 04 AI and agents in PIMWhat vendors have shipped, what is still beta, and what buyers should ask before switching an agent on.
3. 05 Agentic commerce and AI searchAI assistants now shop, compare and sometimes buy. What they need from your product data is surprisingly ordinary.
4. 06 Regulation and product dataMost EU rules for 2027 are about data you already hold, in places you have not yet published it.

Buying

1. 07 Architecture and integrationWhere PIM ends and DAM, MDM, PLM and commerce begin, and why the lines are blurring.
2. 08 Cost, pricing and implementationFew vendors publish prices, nobody publishes an independent benchmark, and the biggest cost is the one nobody counts.

Vendors

1. 09 Vendor landscape and consolidationPlatforms bought the data layer, private equity bought control, and small vendors joined forces.
2. 10 Benelux and DACHOur home market: marketplaces with strict content rules, sector standards with real weight, and most of the vendors in this report.

Buying

1. 11 Choosing a PIMA neutral route from do we need one to a shortlist, and the questions that separate the vendors.

Vendors

1. 12 Vendor profiles40 PIM vendors, alphabetical, facts only.

The last edition did not survive a check, so this one was rebuilt.

We tested 47 figures, dates and named studies in the 2025 report. Eight were still right. Thirteen were out of date. Seven did not match the source they were attached to. One was false. Eighteen could not be traced at all. We cut or replaced the claims we could not defend (our check, 2026).

The market moved while we were checking. AI shopping channels now publish the product fields an assistant needs. Five of the 28 vendors we checked document an MCP server as generally available, so an assistant can read and change the catalogue within the rights the vendor allows (our count, October 2026).

Ownership moved too. Salesforce bought Informatica. Dassault Systèmes, through Centric, bought Contentserv. Cinven took control of Salsify, and Tenzing of Pimcore. Both of those last two deals were still completing when we researched them. This report treats them as closed.

This edition adds what the last one lacked: what vendors have actually shipped in AI, agentic commerce, a regulation timeline, cost and implementation, the Benelux and DACH markets, a buyer checklist, a glossary, and 40 vendor profiles, twelve more than in 2025. We do not rank those vendors. A profile is a set of sourced facts. It is not a recommendation.

Facts in this book were closed on 17 October 2026, unless a page names a later date.

Stephan Spijkers and Chris Jobse
PIMvendors.com

Cited in the file, not printed as codes: R05, R04, R07, R08

Cited in the file, not printed as codes: R06, R04, R08

Cited in the file, not printed as codes: R10, R11, R07, R13

Vendor claim. The number comes from the vendor, or from a survey the vendor commissioned. We have not checked it against the customer's books. Quote it as the vendor's claim, or leave it out.

Unconfirmed. The fact was published, and we could not confirm it on a primary source. It stays visible so the gap is on the page. It is not a finding you should repeat as settled.

High, medium, low. High means a primary source, read recently. Medium means a solid secondary source, or a primary source older than twelve months. Low means one weak source. An inference is labelled Our reading, and it is separate from all three.

Sources are named by author and year. The full entry, with the address and the date we read it, is at the back of the book.

Vendor order is alphabetical. A position in the landscape follows two rules we state there: the segment the vendor sells to, and the step in the value chain it names. Neither is a score.

Dates are written the European way: 18 February 2027.

1. What a PIM isThe product record, and the path it takes from source systems to channels.
2. What a PIM must do, and what it may addThe work every PIM does, and the modules only some of them add.
3. Who uses PIMManufacturers, wholesalers and retailers, as the vendors themselves describe their buyers.

Same facts. A different shape for each channel.

A PIM is the system in which a company manages, enriches and publishes the product information it sells with, for every channel. That is our definition. A 2026 analyst landscape, quoted on a vendor site, describes the same job as a hub that manages, enriches and syndicates product data (novomind 2026).

The record is not the price book and it is not the file store. It is the set of facts a channel needs in order to accept the product, and the selling copy that sits on those facts. ERP still owns stock and the invoice. A DAM still owns the image file and the rights on it. The PIM owns the version of the product that a shop, a marketplace, a catalogue or an assistant will show.

The record moves through three stages. Onboarding brings in what the company already knows: the item from the ERP, the specification from suppliers, the design data from a PLM. Enrichment turns that into something a buyer can use: marketing copy, assets, translations and a classification. Publishing sends a shaped version to the web, to marketplaces, to print and, from this edition, to AI channels. The diagram on this page is that path. Adding AI channels as an endpoint is our reading of where product data now has to land. The three stages themselves are the working model of this report.

A useful record holds six kinds of fact.

• Identifiers, such as a GTIN or an MPN, so a channel can match the product to a product it already knows.
• Technical specifications, the measurements and materials a buyer or a standard asks for.
• Marketing content, the words that sell, kept separate from the specifications so a claim can be checked.
• Assets, or the references to them, so the image the channel shows is the image you approved.
• Logistics data, the weights and packs a carrier and a warehouse need.
• Regulatory data, the fields a law or a retailer now requires before the product can be offered.

Two channels can need different slices of that record. They should not need two different truths. When colour, material or the product name differs between the shop and the marketplace, the PIM has not done the job, however many connectors it lists.

Cited in the file, not printed as codes: R12 S59

Every product that calls itself a PIM should do the same underlying work. The modules around that work are a choice, and vendors make it differently. Chapter 07 shows how the vendors in this report describe their own modules. The split below is ours, so you can read that chart against a fixed list.

Every PIM should do these.

• Keep a central product repository, and reject a record that fails the rules you set.
• Hold a data model: attributes, families, variants and the relations between products.
• Run a workflow with an approval, so a draft is not a live listing.
• Enrich text, including the version a specific channel needs.
• Offer a view per channel: a catalogue, a locale, a feed.
• Import from your own systems and from suppliers.
• Export, and expose an API, so another system can take the record without a spreadsheet in between.

Some PIMs also offer these.

• A DAM, or a link to one.
• Connectors that syndicate to marketplaces and to data pools.
• Translation and localisation.
• A dashboard for data quality.
• A portal where suppliers complete the record.
• Digital catalogues and data sheets.
• Templates for regulation: a digital product passport, ETIM, food labelling.
• Digital shelf analytics.
• An MCP server or an agent that can read the catalogue, and sometimes write to it.

Treat the second list as a set of questions, not as a score. A vendor that stops at the first list can be the right buy. A vendor that advertises the whole of the second list still has to show which items are in the product you would licence, and which are a partner, a beta, or an extra contract.

Cited in the file, not printed as codes: R12

Of the 28 vendors in the 2025 edition, 21 name manufacturers as an audience, 19 name wholesale or distribution, and 23 name retail or brands. Thirteen name all three. The source is each vendor's profile in the PIMvendors directory, read on 10 October 2026, and the confidence is medium. The text is partly the vendor's own positioning and partly our summary of it. It says who the vendor claims to serve. It does not say who pays them.

A handful describe a narrower market. Signifikant and SyncForce name manufacturing only. BetterCommerce and Ergonode name retail and brands only. OmniStock and ProductFlow name wholesale and retail, and not manufacturing. Utopis names manufacturing and wholesale, and not retail. Broad or narrow is a fact about the page they publish. It is not a judgement of the product.

The work itself differs by step, even when the software category is the same.

Manufacturers live in technical data. A buyer, and often a standard such as ETIM, wants the specification to be complete and stable. The selling text is secondary to the attribute that a distributor will map.

Wholesalers live in other people's data, arriving in volume. The hard part is onboarding suppliers, mapping their attributes onto one model, and sending a clean record onward. A portal and an import matter here more than a campaign headline.

Retailers and brands live in the channel. The same product needs a different shape for the shop, the marketplace and the print piece, and the facts underneath those shapes have to stay the same. Content per channel is the job. A missing required field is a rejected listing, not a style choice.

If you are in one of these steps, start with the vendors that name it. Then test the claim. A logo strip is not a customer base, and a sentence that says "for retail" is not a reference.

Cited in the file, not printed as codes: R17

1. Five signals that spreadsheets have run outComplexity, conversion, channels, borders and returns, without the figures we can no longer defend.
2. What poor product data costsThe few figures that survived a check, and the kind of evidence they are.

Five pressures show up in almost every selection project. The 2025 edition attached a statistic to each of them. Those statistics are gone. What remains is the pressure, plus the facts we can still source.

The catalogue gets too complex to pass by hand. Variants, packs, seasons and languages multiply the rows. A spreadsheet can hold them while one person knows which file is current. It stops being a system when two people edit it, or when a launch depends on a column nobody owns. We have no sourced count of how many SKUs tip a company over. The signal is operational: people keep a second list because the first one is no longer trusted.

Conversion depends on the record being complete. A channel that rejects a file, or a shopper who cannot find the specification, is a commercial problem with a data cause. The popular percentages for cart abandonment and conversion uplift did not trace to a source we can print. Chapter 02's next page keeps the only consumer figures that did.

Channels stop accepting free text. In 2025, 56.7% of German online sales went through marketplaces (HDE 2026). bol requires an EAN and declines an upload without one. OTTO requires an EAN on new products. Kaufland requires an EAN, and has required image alt text since June 2025. Zalando publishes detailed attribute rules, but the EAN rule itself was not on the public pages we could read, so we do not state it. From 31 January 2027, Google Merchant Center will enforce a minimum image of 500 by 500 pixels (Google 2026). A PIM earns its place when those rules live on the record, and a publish fails before the channel does.

Borders add fields, not just languages. Since 13 December 2024 the General Product Safety Regulation has required, on distance sales, the manufacturer's name and address, an EU responsible person where the manufacturer is outside the Union, a product identifier, and warnings (European Union 2023). Every channel that offers the product has to show them. A translation memory does not collect a field you never stored.

Returns close the loop. Wrong or thin product information is one reason goods come back. It is not, on the evidence we have, a ranked cause. The next page says what can be said about the cost, and what cannot.

Cited in the file, not printed as codes: R15 S04, R15 S08, R15 S14, R15 S16, R15 S12, R08 S22, R10 S20

Vendor surveys measure a vendor's question. They are not an independent cost of poor PIM data.

Buyers ask what bad product data costs. The honest answer is that almost every widely quoted number failed a source check. Four findings are solid enough to print, and each one carries a limit.

Akeneo reports that 65% of consumers would abandon a purchase when product information is lacking, and that 40% returned a product because the information they saw before buying was inaccurate (Akeneo 2026). Both figures are a vendor claim. We have them from a secondary report of 31 March 2026, not from a method note we read ourselves. They describe shoppers answering a vendor's survey. They do not describe your return rate.

Syndigo, with YouGov, surveyed 8,736 consumers in June 2026 and reports that 80% say inaccurate content damages the brand (Syndigo 2026). That is also a vendor claim. The same survey is described as showing shoppers more likely to buy what an AI recommends, but Syndigo's own page and a secondary report disagree between 25% and 26%. Until the press release is checked, this book does not print that figure.

The US National Retail Federation, with Appriss Retail, reported that US returns in 2023 were USD 743 billion, 14.5% of sales, and 17.6% for online sales (NRF 2023). The source gives no ranking of why goods come back. We do not treat product information as a measured cause of those returns. The year is 2023. Treat it as scale, not as a current rate.

Put the four findings side by side and the pattern is the finding. Consumer pain is documented by companies that sell the remedy. The one independent retail figure does not connect the returns to product content. An independent study that measures the cost of poor product data for PIM was not found.

Count the cost in your own operation. Rejected marketplace feeds, tickets marked "not as described", and the hours spent reconciling two descriptions of one SKU are evidence you can audit. A percentage from a vendor's survey is a prompt to go and look. It is not the look.

Cited in the file, not printed as codes: R05 S2, R05 S28, R09 S9, R14 S37, R14 S38, R14

1. How big is the market?A range for today, and a much wider range for the next decade.
2. Where the market isRegions, company sizes and industries, publisher by publisher.
3. Hard numbers behind the estimatesThe few company and deal figures that can sit beside a market of billions.
4. What the numbers mean for a buyerDirection is usable. A budget taken from a forecast is not.

Nine publishers released forecasts in 2025 and 2026. Their base values for 2025 run from USD 15.7 billion (IMARC) to USD 23.0 billion (Value Market Research). Fortune Business Insights, at USD 5.5 billion, sits apart. Our reading is that Fortune counts a narrower scope.

Mordor's published base is for 2026, at USD 19.95 billion, so it sits in the set and outside that 2025 span. Value Market Research was read from a reseller listing, not from the publisher's own page, so the upper figure in that range is low confidence.

Under the chart, cloud and on-premises do not agree either. Mordor puts cloud at 63.5% of 2025 revenue. Kings puts on-premises at 59.94% in 2024, and Polaris at 54.61% in 2025. All of them say cloud is the fastest-growing deployment, at about 18–19% a year (Mordor 2026; Kings 2026; Polaris 2026).

Cited in the file, not printed as codes: R06 S1, R06 S2, R06 S3, R06 S4, R06 S5, R06 S6, R06 S7, R06 S23, R06 S24

Europe is under-measured. That is a finding, not a missing footnote.

North America is the largest region in most of these reports, and the share depends on who is counting. Mordor puts it at 39.8% of 2025 revenue. Precedence puts it at 38% for 2025. Kings puts it at 34.09% for 2024. Fortune puts it at 31%, on a base so far below the others that the share is low confidence (Mordor 2026; Precedence 2026; Kings 2026; Fortune 2026).

Polaris is the exception. It names Asia Pacific first, at 35.47% in 2025 (Polaris 2026). Asia Pacific is also the region most publishers call the fastest growing: 20.1% a year at Mordor, 15.06% a year at Kings.

Europe is the gap. Fortune is the only publisher in this set that gives Europe a share, 29%, and Fortune is the low-confidence outlier. Everyone else describes Europe in words or leaves it out. If you need a European number for a board, you do not have one that this report will stand behind. The Benelux and DACH chapter later in the book uses retail and marketplace facts instead of a share of a global software market.

Company size is more consistent, and still not identical. Large enterprises hold 68.8% of the market in Mordor's 2025 cut, 53.27% in Polaris, and 54.72% in Fortune's 2026 cut. Mordor says smaller firms are growing fastest, at 18.43% a year (Mordor 2026; Polaris 2026; Fortune 2026).

By industry, Mordor names retail and e-commerce as the largest slice, at 33.2%, and IT and telecom as the fastest growing, at 19.5% a year (Mordor 2026). Other publishers do not line up on the same industry list closely enough to merge. We do not merge them.

Read a chart of regions as a set of publisher decisions about scope. A company selling in Germany is not 29% of a global total, and a forecast will not tell it so.

Cited in the file, not printed as codes: R06 S2, R06 S3, R06 S4, R06 S5, R06 S6

A forecast of tens of billions needs something to stand on. The company figures we can verify are smaller, and they are worth more because of that.

Stibo Systems reported revenue of DKK 1.31 billion for 2025–26, up 6.0%, with software-as-a-service revenue up 17.0% and an EBITDA margin of 18.1% (Stibo Systems 2026). The company sells multidomain master data as well as PIM. The figure is the company's own.

Dassault Systèmes, through Centric, bought Contentserv at an enterprise value of EUR 220 million (Dassault Systèmes 2025). Salesforce bought Informatica for about USD 8 billion in equity value. The deal closed on 18 November 2025. It is a master-data and data-management transaction, not a purchase of a PIM product on its own (Salesforce 2025).

Syndigo described the company that resulted from combining with 1WorldSync as worth more than USD 3.5 billion (Syndigo 2025). That is the company's framing of its own combination, not a price paid for a PIM module.

ComCap, an advisory firm, counted 17 mergers and acquisitions and 21 private placements in this space from January 2024 to March 2026. The median private placement was USD 3.3 million (ComCap 2026). The median is a useful scale for the long tail. It is not a typical price for the platforms named above.

Set Stibo's revenue next to a market reported at about USD 16–21 billion and the gap is the point. Our reading is that the forecasts include implementation, content services and adjacent software, not only PIM licences. A services-heavy definition can be the right description of what a buyer spends. It is the wrong description of what a vendor invoices for the product.

Current recurring revenue for Akeneo, Pimcore, inriver and Salsify was not found in a source we will print. Older aggregator figures stay out.

Cited in the file, not printed as codes: R06 S17, R06 S15, R07 S8, R06 S13, R06 S25, R07 S11, R07 S12, R06 S16, R07 S14, R06 S19

Direction is usable. A number copied into a business case is not.

A growing category means the shortlist will stay crowded, and that owners will keep changing. Chapter 09 records who bought whom. Choice and consolidation are the same fact seen from the buyer's side of the table: more logos, fewer independent ones.

Cloud is the deployment publishers say is growing fastest. That matches how vendors now package the product. Our reading of the pricing and release evidence is that new vendors and new releases are software-as-a-service first (see chapters 04 and 08). An on-premises option still exists at some of them. It is no longer the story the forecasts are telling.

None of the nine publishers gives a cost per SKU or a cost per user. A forecast cannot be turned into a licence budget by dividing it by a count of companies, because the publishers do not share a definition of the company or of the spend. Use the reports to answer a board that asks whether this category is growing. Answer the budget question with quotes, with the internal time you will spend, and with the payback range in chapter 08, which is our guidance and not one of these forecasts.

The public article on pimreport.com about market size in 2026 prints none of these billions. It asks the reader to check what a publisher counted. This chapter is the longer version of that advice, with the ranges filled in and the outliers left visible.

Cited in the file, not printed as codes: R06, R04, R13

1. What is documented as availableGenerative copy and translation lead the documentation. Agents and MCP are fewer.
2. A maturity ladderFive stages, from no embedded AI to a closed loop nobody documents yet.
3. Surveys and statisticsVendor surveys and official figures describe different populations.
4. What to ask before a writeApproval, model, residency, cost and permissions.
5. The AI Act and product contentWhat applies to product text, and what is consumer law instead.

"Generally available" means the vendor's own documentation, help centre, release notes or a dated release says customers can use it. Beta means the vendor labels it so. Announced means marketing only. The count is the 28 vendors of the 2025 edition, read on 13 October 2026. The twelve vendors added in this edition are not in the chart.

Cited in the file, not printed as codes: R04

Where is your team?

We use five stages. They are our analysis, not a standard.

1. Unevidenced. No embedded AI in the documentation. People paste text into a separate chat tool. There is no policy.
2. Assisted. The product can generate, translate, extract or map, and a person edits each result.
3. Governed. A write waits for approval. The team can choose the model, and prompts and decisions leave an audit trail.
4. Agentic. An agent or an MCP server works on the catalogue inside permissions, with a gate on writes.
5. Closed loop. The agent acts under a policy, and quality or commercial results are measured against a human baseline. No vendor documents this stage.

Counted on generally available documentation only, the 28 vendors sit at 21, 2, 1, 4 and 0. If beta counts, the line is 19, 2, 0, 7 and 0 (our classification, 2026).

A separate, vendor-hosted index of 405 firms found that 13.6% report end-to-end automation with an audit trail (inriver 2025). That is a vendor claim about buyers, not a count of products. Use the ladder to mark where your team is. Do not use it to rank the vendors in chapter 12.

Cited in the file, not printed as codes: R09 S25, R09 S4

inriver asked 317 executives in the first quarter of 2025. It reports that 97% are past experimentation and that 83% have AI embedded across systems (inriver 2025). The same firm's maturity index, 405 companies, found that 117 claimed full agentic readiness and that none of those 117 met the index's own benchmark. It also reports that 33% keep at least 85% of SKUs ready to publish (inriver 2025). Both are vendor claims.

Akeneo asked 1,000 IT decision makers in the United States and Europe in September 2026. It reports that 95% are confident they are ready for AI-driven commerce, and that 89% spend more than a quarter of each AI project on preparing data (Akeneo 2026). That is a vendor claim. Readiness and data preparation are both in the same survey. The second number is the one that matches what implementation teams describe.

Eurostat, for 2025, reports that 20.0% of EU firms with ten or more staff use any AI, and that 8.8% generate text (Eurostat 2025). A secondary source puts the Netherlands at 33.2% and Germany at 26.0%. Those two country figures are unconfirmed until they are read off the Eurostat table.

The populations do not match, and that is why the headlines do not match. Executives at PIM buyers say AI is already embedded. The official series says most European firms do not use it at all.

What blocks the work, in the order the evidence supports, is data quality and preparation, then trust. inriver's executives report that 90% still see occasional accuracy problems, and that 68% measure trust with indirect indicators (inriver 2025). Cost, skills and the law come after. The legal share is not quantified in these surveys.

Cited in the file, not printed as codes: R09 S1, R09 S2, R09 S3, R09 S4, R09 S5, R09 S7, R09 S8

These are documented facts, not a ranking. Each one comes from the vendor's own pages.

Akeneo documents Ask Ziggy as generally available since May 2026. Agentic Ziggy is in beta, with an approval plan before a change and enrichment that can be reverted. An MCP server is documented. Whether a given package includes it depends on the edition (Akeneo 2026).

Bluestone documents an MCP server as generally available, including permitted write and destructive operations. Its agent asks for confirmation before a write. The documented model providers are Anthropic, Google and OpenAI (Bluestone 2026).

inriver's spring 2026 release documents Enhance and Expression agents and a visual workflow builder with agent actions. The summer 2026 release documents an Enrich Assistant and a further MCP server (inriver 2026).

Sales Layer documents an AI suite and an MCP server on 11 December 2025, an agents knowledge base on 30 July 2026, and Anthropic as a model provider on 13 August 2026 (Sales Layer 2025; Sales Layer 2026).

Stibo Systems documents an MCP server as generally available in release 2026.2 on 16 June 2026. CoPilot chat is in limited availability in 2026.3. AgentWorkx was announced on 29 September 2026 (Stibo Systems 2026).

Pimcore's Agent Bundle, in release 2026.2, is beta. It documents propose, review and approve for every write (Pimcore 2026).

Put these questions to any vendor, including the six above.

• Is the feature generally available or beta, in a document rather than a slide?
• Which model runs it, and in which country is the data processed?
• Can you bring your own model?
• Does every write need a person to approve it, and is the write logged and reversible?
• Which AI features sit in the base licence, and which are credits or tokens on top?
• Does the MCP server respect the same field permissions as a human user?

Cited in the file, not printed as codes: R04

This page is not legal advice.

Regulation (EU) 2026/1744 entered into force on 27 July 2026. High-risk duties now apply from 2 December 2027 for Annex III and from 2 August 2028 for Annex I (European Union 2026).

Article 50, on transparency, has applied since 2 August 2026. Systems already on the market have until 2 December 2026 for machine-readable marking. That grace period is low confidence in our notes (R09).

Our reading, and it is an inference, is that generating, translating, classifying and enriching product data does not fall in the Annex III high-risk categories, so those uses are typically outside the high-risk regime. AI literacy under Article 4 still applies to anyone deploying AI. A secondary account says the duty was softened to taking measures that support literacy. That wording is low confidence until it is read on EUR-Lex.

The rules that change a product page sooner are consumer law. From 27 September 2026, a generic environmental claim without proof is banned (Directive (EU) 2024/825). "Eco", "green" and "climate friendly" need an evidence link or they come off the record.

The new Product Liability Directive covers software and AI for products placed on the market after 9 December 2026 (European Union 2026). A generated description is not outside that only because a model wrote it.

Our reading on copyright: text that is wholly generated is unlikely to be protected in the European Union or the United States, so copy that no person edited can be copied by someone else (inference). Keep a human edit, and a record of who approved it, if the wording has to stay yours.

Cited in the file, not printed as codes: R10 S14, R05 S37, R09 S17, R09 S16, R09 S18, R10 S40, R09 S19, R09 S20, R09 S21

1. Where it stands in EuropeDiscovery is wider than checkout. European checkout dates were not found.
2. TrafficUS referrals are growing from a base under 1% of visits.
3. What the channels requireThe same core fields, plus identifiers that decide a match.
4. A checklist for 2027Eleven jobs for the person who owns the catalogue.

Adyen reports that 25% of Dutch consumers used an AI assistant while shopping in the past year (Adyen 2026). That is a payments firm's survey, medium confidence. It is about research, not about an assistant paying.

Cited in the file, not printed as codes: R08 S48

Bain, in November 2025, put AI referrals under 1% of total US retail traffic, and as high as 25% of referrals at some retailers (Bain 2025). Similarweb reports that 11.4% of shopper journeys involve AI, and that most of that influence does not show up as a click (Similarweb 2026). Both are US-weighted. An EU or NL share of traffic was not found.

Cited in the file, not printed as codes: R08 S42, R08 S43, R08 S44, R08 S45

OpenAI's product feed, Google Merchant Center, Google's Universal Commerce Protocol and schema.org ask for the same core. MCP itself defines no commerce attributes (MCP specification 2026). Google says AI Overviews and AI Mode need no special markup beyond ordinary structured data (Google 2026).

A 2023–24 paper on generative engine optimisation reported gains of up to 40% on general queries. It is not about product pages, and it is not a reason to rewrite a catalogue for an assistant.

Cited in the file, not printed as codes: R08 S26, R08 S49, R08 S27

1. 1Complete ID, title, description, brand, image, price and availability on every SKU you sell.
2. 2Fix GTIN check digits and MPN first. Do not invent an identifier. OpenAI checks the digit.
3. 3Model variants with a group ID, colour, size, gender and age group where the channel requires them.
4. 4Meet image rules before 31 January 2027. Google Merchant Center then enforces 500 by 500 pixels.
5. 5Set your own refresh rule for price and availability. No specification states an interval.
6. 6Publish specifications, dimensions, materials and compatibility as facts, not as adjectives.
7. 7Map one PIM record to a feed per AI surface you actually sell on.
8. 8Ask whether MCP or API read access is generally available, and start with read-only permissions.
9. 9Skip GEO tricks. Keep pages indexable. Google says ordinary structured data is enough.
10. 10Audit claims, prices and safety data. Product liability for software applies after 9 December 2026.
11. 11Measure AI visibility in referrals and mentions. Most of the influence never appears as a click.

Cited in the file, not printed as codes: R08

1. The timelineWhat is in force, what is dated, and what is still a proposal.
2. Live duties and open actsBatteries have a date. Most product passports do not.
3. Must act in 2027Ten jobs, each with a date.
4. Digital product passportsThe carrier is not a mandate for one barcode syntax.
5. Sunrise 2027A voluntary checkout target, plus ten preparation steps.

1. 13 December 2024GPSR applies.
2. 28 June 2025European Accessibility Act.
3. 1 July 2026CLP online labels.
4. 19 July 2026Unsold-goods destruction ban for large companies.
5. 20 July 2026Digital product passport registry is live.
6. 2 August 2026AI Act Article 50 transparency.
7. 12 August 2026Packaging regulation applies.
8. 27 September 2026Empowering Consumers Directive.
9. 9 December 2026Product Liability Directive.
10. 30 December 2026EUDR for large and medium operators. Micro and small: 30 June 2027.
11. 1 February 2027CBAM certificate sales.
12. 18 February 2027Battery QR code and battery passport.
13. 2 December 2027AI Act high-risk, Annex III. Annex I follows on 2 August 2028.
14. 12 August 2028PPWR labelling, no earlier than this date.
15. 26 July 2029CSDDD.
16. 1 January 2030PPWR recyclability, no earlier than this date.
17. 19 July 2030Unsold-goods ban for medium companies.

Cited in the file, not printed as codes: R10

These duties are live. Distance sales under the General Product Safety Regulation must show the manufacturer, an EU responsible person where the manufacturer is outside the Union, an identifier and warnings (European Union 2023). The European Accessibility Act applies, with an exemption for microenterprise services (European Union 2025). Generic environmental claims without proof are restricted from 27 September 2026 (European Union 2024). CLP pictograms, the signal word and hazard statements belong online for chemical products (European Union 2026). An energy label sits in EPREL and is shown with the price (European Union 2017).

Batteries have the firmest 2027 date. From 18 February 2027 every battery needs a QR code. Electric-vehicle, light-transport and industrial batteries over 2 kWh need a battery passport (European Union 2023). The label date is 18 August 2026 or 18 months after a labelling act, whichever is later. No adopted labelling act was found, so the label is probably later.

No delegated act yet sets the passport attributes for a product group. The working plan points to iron and steel in the fourth quarter of 2026, textiles, aluminium and tyres in the third or fourth quarter of 2027, furniture in 2028 and mattresses in 2029, each with at least 18 months of transition (European Commission 2025). Those are plans, not obligations.

Two regimes are smaller than the 2025 edition implied. CSRD, on a law-firm reading of Directive (EU) 2026/470 that still needs the Official Journal, covers companies with more than 1,000 employees and more than EUR 450 million turnover. CSDDD covers companies with more than 5,000 employees and more than EUR 1.5 billion turnover, from 26 July 2029. Confirm both thresholds on EUR-Lex before you repeat them.

Watch, and do not implement as law: a June 2026 proposal to replace the printed energy label with a QR code and an EPREL number; a pending extension of CBAM downstream; and a Digital Fairness Act expected in the fourth quarter of 2026.

Cited in the file, not printed as codes: R10 S20, R10 S12, R10 S40, R10 S27, R10 S29, R10 S7, R10 S37, R10 S3, R10 S16, R10 S32, R10 S34, R08 S51

1. 1From 18 February 2027, give every battery SKU a QR code. EV, light-transport and industrial batteries over 2 kWh need a passport link.
2. 2From 30 December 2026, EUDR goods need an HS code, country of production and a due-diligence reference per SKU.
3. 3From 27 September 2026, audit eco, green, climate-friendly and biodegradable claims. Keep an evidence link.
4. 4CLP online labels apply. From 1 January 2027 the next chemical step needs pictograms, a signal word and hazard statements.
5. 5Build packaging data per SKU: material, weight, recyclability, reuse. The label itself is no earlier than August 2028.
6. 6Complete GPSR and accessibility fields in every channel: manufacturer, EU responsible person, identifiers, warnings.
7. 7Hold the EPREL number, the label and the information sheet, and show the label with the price.
8. 8Design the product ID, operator ID and facility ID now. Attribute lists for other passports come later.
9. 9Apparel and footwear: the unsold-goods disclosure format applies from February 2027. Keep stock by category.
10. 10Record which texts and images are AI-generated and who approved them. Marking for systems already on the market runs to 2 December 2026.

Cited in the file, not printed as codes: R10

Do not write that a digital product passport requires GS1 Digital Link.

The Ecodesign for Sustainable Products Regulation has been in force since 18 July 2024 (European Union 2024). The digital product passport registry went live on 20 July 2026 (European Commission 2026). Implementing Decision (EU) 2026/1736, of 14 July 2026, cites the harmonised standards EN 18216 and EN 18219 to EN 18223 (European Union 2026).

What is not decided is the attribute list for each product group. Those lists sit in delegated acts that have not been adopted. The only passport legally required in 2027 is the battery passport.

The data carrier is a URL that resolves and is globally unique. EN 18219 allows more than one syntax. GS1 Digital Link is one permitted option. Some readers of the standard recommend it. The standard does not mandate it. A page that says the passport requires GS1 Digital Link is wrong.

The practical step is ours, and it is an inference from the texts above. Give each product a unique identifier, name the operator and name the facility. When a delegated act arrives, those three identifiers are what the new attributes will hang from.

Cited in the file, not printed as codes: R05 S16, R10 S3, R11 S15, R05 S35, R05 S36, R11 S33

Guideline 1.1.0, December 2025, is the text. From 2028 the brand owner chooses the carrier. Dual marking continues until point-of-sale readiness is high, in the range of 90–95% (GS1 2025). No current measurement of that readiness was found. The last widely cited figure is from 2021 and is stale, so it is not printed.

In the Netherlands, GS1 and Albert Heijn expect Dutch checkouts to be ready by the end of 2027. GS1 Nederland lists more than 30 Dutch retailers as involved. Involvement is not the same as a scanner that works. No statement was found for Jumbo, Lidl in the Netherlands, or the Ahold Delhaize group.

Related standards, separate from the checkout target: GS1 Digital Link URI syntax 1.7.0 and Resolver 1.2.1, August 2026, medium confidence. GDSN 3.1.35 went live on 16 May 2026. ETIM 11.0 is final on 1 December 2026.

Cited in the file, not printed as codes: R11 S3, R11 S19, R11 S27, R11 S6, R11 S25, R11 S32, R11 S28, R11 S30

1. 1Confirm the target is voluntary and that 1D codes stay valid. Ask each key retailer what it will require, and when.
2. 2Audit GTIN quality at each packaging level. The same GTIN has to sit in every code that will be scanned.
3. 3Choose the carrier per product family: QR with Digital Link, Data Matrix with Digital Link, or GS1 DataMatrix.
4. 4Plan dual marking until those retailers confirm they accept the 2D code.
5. 5Our recommendation: store carrier, URI or domain, link set and landing-page status as PIM attributes.
6. 6Name who owns lot, expiry and serial, and how the ERP and the printer talk to the PIM.
7. 7Choose a resolver and check it against Resolver 1.2.1, including the default link and the description file.
8. 8Prepare a destination page per audience and language, and a fallback when the link fails.
9. 9Put code generation, print verification and change control into the artwork process.
10. 10Decide on purpose whether the passport link and the Sunrise link are the same URI. EN 18219 allows more than one syntax.

Cited in the file, not printed as codes: R11

1. Four labelsPIM, DAM, MDM and PXM, and the three points where the definitions split.
2. What vendors name as nativeSyndication and DAM are common. MDM, PLM and CPQ are not.
3. Suite or separate toolsWhich situation leans which way.
4. The 2027 mapThe old neighbours, plus agents, shelf analytics and pricing.

The definitions disagree in three places.

PIM can mean the content hub for channels, or the product master for the company. A vendor that says "we do MDM" may mean a deep multi-domain system or a product record with a workflow. Ask which domains have a golden record.

PXM may be a separate category. IDC treats PXM and PIM+ as different evaluation groups (IDC 2025). The public Forrester material we could read does not define PXM as its own market. Our compare pages, and this report, treat PXM as PIM plus a claim to test.

MDM, when the word is used strictly, covers more than product. Product content for selling can live in a PIM while customer and supplier masters live elsewhere. Mid-market firms often let the PIM act as the commerce product master and skip a separate MDM. That can be right. It is not the same as buying both.

Cited in the file, not printed as codes: R12 S59, R12 S60, R12 S66, R12 S58

The chart counts modules the vendor names as native. PIM is 28 of 28. Syndication is 22. DAM is 17. MDM is 5. Feed management is 4. Commerce is 5. DXP or CMS is 4. PLM is 2. CPQ is 1. The twelve vendors added in this edition are not in these counts.

Cited in the file, not printed as codes: R12

Our reading of the deals in chapter 09: when a platform acquires the PIM, add one question the table does not show. Ask whether the PIM product is still positioned as strategic, in a document, not in a keynote.

Cited in the file, not printed as codes: R12 S61, R12 S65, R07

The 2025 map still holds. The ERP supplies operational facts. The PIM curates what you sell and what you must show. PLM hands over when a design becomes a commercial product. DAM is linked, not copied. MDM, where you have it, governs the shared masters. Commerce and the DXP exchange data in both directions.

Three nodes were missing from that drawing.

AI and agent channels take a feed. Shopify says its catalogue syndicates merchant data to AI surfaces. That is Shopify's claim. OpenAI and Google publish feed specifications of their own, covered in chapter 05.

Digital shelf analytics reads how the product appears on a retailer site. Several PIM vendors now sell or partner for that view. It is not the same module as the record itself.

Pricing sat outside the PIM until recently. Akeneo announced the purchase of PricingHUB on 11 June 2026, as a dedicated business unit (Akeneo 2026). A price engine and a product record can share a vendor. They are still different jobs.

Around that map, the suites have moved. Salesforce owns Informatica's product and master-data software. Adobe's commerce tooling can ingest from an external PIM. SAP's master-data governance sits on the material master. Our reading: none of those arrangements replaces a dedicated PIM on the evidence found. Each still expects a product record to exist somewhere.

Cited in the file, not printed as codes: R12 S65, R12 S44, R12 S45, R12 S46, R12 S47, R12 S48, R07 S1

1. How vendors priceSix models, and the extras that do not show on the first quote.
2. Published entry pricesA short list. The tiers do not buy the same thing.
3. How long a rollout takesVendor guides and a few cases. Most cases give no duration.
4. Total cost and paybackInternal time is the missing line. Payback here is our guidance.

Vendors in this set price in six ways. The examples are illustrations of the model, not a shortlist.

Pimcore sells editions, and states that users and products are not capped on those editions. Plytix, Ergonode and Katana PIM tier by SKU or product count. Sales Layer and Quable's Standard plan sell seats with a SKU cap. Bluestone quotes on usage. Struct tiers by users and states no SKU limit. AtroPIM, Pimcore Community, Ergonode Free and UnoPIM combine an open-source or free core with paid support or modules.

Whatever the meter, ask five questions before you compare two quotes.

• Which AI features are in the base price, and which are credits or tokens?
• Which modules are separate contracts, in particular a supplier portal and analytics?
• What happens to the price at renewal?
• What is the minimum term?
• Is a fee charged per record type, or per concurrent user, on top of the tier you were shown?

Two quotes with the same annual number can buy different volumes, different modules and a different renewal. Chapter 08's next page shows the few prices that are public, and why they still do not compare.

Cited in the file, not printed as codes: R13

Akeneo, Bluestone, Sales Layer, Salsify and Quable do not publish a list price. Catsy, outside this profile set, is also quote only. Do not fill the gap with a competitor's estimate.

Cited in the file, not printed as codes: R13 S06, R13 S07, R13 S08, R13 S14, R13 S15, R13 S18, R13 S22

Plytix's guide says 4 to 12 weeks for a smaller first channel, and 6 to 18 months for an enterprise or a catalogue above 50,000 SKUs (Plytix 2026). inriver's guide says 3 to 6 months for a smaller firm and up to 12 months for an enterprise (inriver 2026). Neither guide cites a sample.

Three published cases give a duration, each as a vendor or partner claim. Marc O'Polo: 5 weeks to go live, with an SAP integration. Carpeto: 4 months, on more than 20,000 SKUs. Zur Rose Pharma: 5 months to a first go-live and 11 months for the whole project. Most case studies we opened state no duration and no cost.

Cited in the file, not printed as codes: R13 S36, R13 S37, R13 S38, R13 S39, R13 S40, R13 S41

Payback of 12 to 24 months is PIMvendors guidance. It is not a study.

A full cost has eight parts: the licence or subscription, implementation, integrations, migration and cleansing, training, support and maintenance, running costs, and the time of your own staff. The last part has no public benchmark. It is the cost buyers leave out of the spreadsheet.

Vendors do publish ratios, and they are claims. Bluestone says 20–40% of the first-year budget goes to migration, onboarding, configuration and integration, and 10–15% to training and process redesign (Bluestone 2026). inriver says enterprise implementation costs at least one annual licence, and that on-premises maintenance is 15–25% of the licence a year (inriver 2026). A PIMvendors knowledge-base note uses that same 15–25% maintenance band. It is our note, not an independent audit.

Payback, in our guidance, takes 12 to 24 months. That replaces earlier ranges we have used. It is experience, not a measured sample. Vendor-commissioned studies claim faster payback. A Forrester Total Economic Impact study commissioned by Salsify reports 7 months (Forrester 2024, commissioned). Label it as commissioned if you cite it. Do not average it with our range.

For context only, and it is not about PIM: Flyvbjerg and Budzier, in 2011, found an average IT cost overrun of 27%, and about one project in six near a 200% overrun. The paper is old. No PIM-specific failure rate exists. Do not invent one, and do not reuse any "70% of projects fail" line. Those lines did not survive the check of the 2025 edition.

The cost calculator on pimreport.com is not used in this chapter. Its unit of work is still undefined, and its sample profiles sat above the public prices. Until that unit is explained, the calculator stays off the page.

Cited in the file, not printed as codes: R13 S28, R13 S30, R13 S04, R13 S48, R13 S51

1. Who bought whomFour patterns from 2023 to 2026.
2. What ownership changes for a buyerThree groups of risk, and two clauses to negotiate.
3. Where the 40 sitSegment and value chain only. Not a ranking.
4. Who the public evaluations nameInclusion where a public page confirms it. No scores.

Four patterns cover the deals we could date.

Platforms bought the data layer. Salesforce announced Informatica on 27 May 2025 and closed on 18 November 2025, at about USD 8 billion of equity value. SAP announced Reltio on 27 March 2026 and closed on 7 May 2026. Centric, part of Dassault Systèmes, agreed to buy Contentserv for an enterprise value of EUR 220 million on 25 February 2025. Epicor bought KYKLO on 14 June 2024.

PIM vendors widened what they sell. Akeneo bought Unifai on 19 September 2023 and announced PricingHUB on 11 June 2026. Syndigo bought 1WorldSync on 3 September 2025 and Taggstar on 10 March 2026.

Private equity took control. Boyum IT, backed by Volpi, bought Perfion in 2023. Cinven announced control of Salsify on 22 July 2026. Tenzing announced control of Pimcore on 27 August 2026. This report treats those two 2026 deals as closed.

Smaller vendors were absorbed. AGENA3000 bought SRC. The working date is 15 May 2024. Katana PIM bought Beeyond on 6 February 2025. Venzee closed the purchase of Jasper on 31 December 2025. XSARUS bought Calago in July 2025.

The stated reason, across the platform announcements, is data ready for AI.

Cited in the file, not printed as codes: R07

Our reading is that the risk is different in each group. Inside a platform, the PIM may become one module among many. Ask whether it still has its own roadmap. Inside a new fund, the risk is price and hold period. Ask what the renewal looks like. Inside a roll-up, the risk is that the product you saw in 2025 now shares a name, a support desk or a codebase with something else. Ask what stayed.

None of this is a reason to prefer one group. It is a reason to put two clauses in the contract: what happens if control changes again, and what the price may do at renewal.

Cited in the file, not printed as codes: R07

The landscape figure is not drawn in this preview. The words under it are.

The position of a logo follows two rules only. The horizontal place is the segment: entry-level, mid-market, or enterprise, taken from the 2025 edition or from the profile where the vendor is new. The band is the value-chain step the vendor names: manufacturing, wholesale and distribution, or retail and brands. A vendor that names more than one step appears in more than one band.

The figure is not a ranking. Distance from the centre means nothing. Absence from a band means the vendor did not name that audience, not that it cannot serve it.

Cited in the file, not printed as codes: R17

Absence from an analyst list says nothing about quality.

ISG's Buyers Guide for PIM platforms in 2026 covers 21 providers in a free four-page summary (ISG 2026). Info-Tech's Data Quadrant for PIM, July 2026, covers 13 products and is free (Info-Tech 2026). Forrester's PIM landscape of the third quarter of 2026 lists 23 vendors. Of those names, only novomind is confirmed on a page we could read. The others are gated. IDC's 2025 MarketScape is gated. A full inclusion grid for all 40 vendors is not printed, because we will not fill a blank cell with a guess.

Vendors from the 2025 edition that a public 2026 summary we could read does name, as a group and not as a ranking: Akeneo, Bluestone PIM, censhare, Centric, inriver, novomind, Pimcore, Sales Layer, Stibo Systems, Syndigo and Viamedici. Perfion's owner, Boyum IT, appears on an emerging-provider list. That link is our inference. Among vendors added in this edition, Salsify and Plytix are named in those public ISG materials.

No score, no medal and no quadrant is printed. A vendor that is missing from this paragraph may be in a gated report, or in none. Both are possible. Neither is a finding about the product.

Cited in the file, not printed as codes: R14 S07, R14 S10, R03 S22, R03 S23, R03 S24

1. Online retailCompare growth, not the size of the three markets.
2. Marketplace rulesbol, Zalando, OTTO and Kaufland. Amazon was not researched.
3. StandardsGS1, ETIM, 2BA, ECLASS and BMEcat.
4. Vendors, partners, eventsWhere the 2025 set is based, and three dates for 2027.

Germany's online sales of goods were EUR 92.3 billion in 2025, up 3.9%, excluding services and VAT. The HDE forecast for 2026 is EUR 96.3 billion (HDE 2026). The Netherlands recorded EUR 35.7 billion, down 1%, and that series includes services. Product sales in the Netherlands grew 2% (Thuiswinkel.org / GfK 2026). Belgium recorded EUR 18.3 billion, up 5.4% (Becom 2026). Compare the growth rates. Do not add the totals.

Marketplaces carried 56.7% of German online sales in 2025 (HDE 2026). HDE also reports that 35% of Germans use AI when researching a purchase (HDE 2026). A third-party compilation puts bol at about 13.5 million customers and 62,000 sellers. That pair of figures is low confidence and needs a primary source before it is repeated.

Cited in the file, not printed as codes: R15 S04, R15 S05, R15 S01, R15 S06, R15 S11

Amazon was not in this research. Do not copy another marketplace's rules onto Amazon and call them current.

Cited in the file, not printed as codes: R15 S08, R15 S09, R15 S10, R15 S12, R15 S13, R15 S14, R15 S16

GS1 Nederland, GS1 Belgilux and GS1 Germany run the identifier and the data-pool services, and they publish support for digital product passports. The clearest enforced scheme we found is GS1 Germany's data-quality programme for food and near-food, which travels through GDSN. A record that fails it does not travel.

ETIM is the classification technical wholesalers use. It has 16 national organisations. The headquarters is in Margraten, in the Netherlands. ETIM 11.0 is final on 1 December 2026 (ETIM 2026). 2BA, for installation and construction, sits on ETIM. Calling 2BA mandatory in practice is our inference from how the sector buys. It is not a statute. Label it that way if you repeat it.

ECLASS 16.0 and BMEcat are the other two formats a German industrial buyer may require. BMEcat is the exchange file. ECLASS is the classification inside it. A PIM that cannot map to either will not pass a technical wholesale onboarding, whatever its commerce connectors look like.

Cited in the file, not printed as codes: R15 S23, R15 S26, R15 S30, R11 S30

Of the 28 vendors in the 2025 edition, 17 are headquartered in the Benelux or in DACH: 10 in the Netherlands, 5 in Germany, 1 in Belgium and 1 in Austria (our count from the 2025 edition and the vendor sites, 2026). Bluestone is inside the Dutch ten because the 2025 edition placed it there. Its website suggests Norway. That conflict is unresolved, so do not treat the country as settled.

Others keep an office without moving the headquarters. Akeneo lists Düsseldorf and Amsterdam. Stibo Systems lists Hamburg, Utrecht and Geneva. inriver lists Munich and Amsterdam. Syndigo lists Cologne.

Partner pages for Pimcore, Akeneo, Stibo Systems, Bluestone and SyncForce show local firms. Those directories are the vendor's own list. They are not complete, and a logo is not a commitment of hours.

Three 2027 events are confirmed. E-Commerce Berlin Expo is on 17 and 18 February. Webwinkel Vakdagen is on 7 and 8 April in Utrecht. dmexco is on 22 and 23 September in Cologne. A 5 to 8 April date for Hannover Messe comes from trade press only. Confirm it on the organiser's site before you print it as fact.

Cited in the file, not printed as codes: R15 S57, R15 S48, R15 S49, R15 S50, R15 S51

1. A decision treeSix branches. The end of each branch is a type of system, never a vendor name.
2. Twenty-five questionsEach one can be answered yes, no, or with a document.
3. Why projects stallNo failure rate. The causes we can actually cite.
4. Data-quality measuresDefine them before go-live. Do not borrow a benchmark.

Start with the pain. If channels disagree and the content drifts, you are looking at a PIM. If the same customer, supplier and product facts must survive across the company, look at MDM first. If the files and the rights are the chaos, look at a DAM and link a PIM to it.

Then ask who owns the work. One team, and a small IT group, can carry a suite that includes media and publishing. A DAM, an ERP or a commerce engine that is already in place points to a separate PIM with connectors.

Then the catalogue. Simple products, under about 5,000 SKUs, and a short attribute list, can live in a lighter system. Variants, configured products or engineered products mean you check who owns PLM and CPQ before you pick the PIM.

Then the channels. Marketplace-heavy sales need native syndication, and a feed if the PIM does not name one. Only 4 of the 28 vendors in the 2025 edition name feed management as native. AI channels need a feed you control and a written answer on MCP, from chapters 04 and 05.

Then the regulation that actually applies to your products. Batteries, textiles, food and energy labels need DPP, ETIM or GDSN support in writing, not as a slogan. Chapter 06 is the date list.

Then the money. A published tier and a quote are different kinds of number. Take a three-year view and include your own staff time. Chapter 08 is the method. Payback of 12 to 24 months is our guidance, not a result from this tree.

The end of every branch is a type: a content PIM, an MDM-first sequence, a DAM plus a PIM, a suite, a separate PIM, a feed tool beside the PIM. It is never a vendor.

Cited in the file, not printed as codes: R14, R12, R13

Data model

1. Can you show a product family, variants and relations in our data, not in a sample?
2. Can a required attribute block publication for one channel and not for another?
3. Can we import a supplier file and see the mapping, including what was rejected?
4. Is there a completeness score per channel, and can we change the rules?

Workflows and governance

1. Who can approve a change, and is the approval stored with the user and the time?
2. Can we revert one attribute without reverting the whole product?
3. Can a supplier complete fields in a portal without seeing the rest of the record?

AI and agents

1. Is each AI feature generally available or beta, in a document?
2. Which model runs it, and where is the data processed?
3. Can we bring our own model?
4. Does every write need approval, and is it logged and reversible?
5. Does an MCP server use the same field permissions as a person?

Channels and syndication

1. Which marketplaces have a native connector, and which are a partner?
2. Can you show a feed for the AI surfaces we named, mapped from one record?
3. What happens to a product when a required marketplace field is empty?

Regulation and standards

1. Show a battery passport, an ETIM class or a GDSN message on a product we choose.
2. Where do GPSR manufacturer, responsible person and warnings live on the record?
3. Can you store a resolver URI, a carrier type and a landing-page status?

Integration and architecture

1. What is the handover from our PLM or ERP, in a diagram of our systems?
2. Does the PIM link to our DAM by reference, or does it store a second file?
3. Which of MDM, commerce and CPQ are in this product, and which are not?

Commercials

1. Is there a list price, or only a quote? What unit does the price meter?
2. Which AI features and modules cost extra, as credits, tokens or a separate contract?
3. What is the minimum term, and what can the price do at renewal?
4. What happens to price and continuity if the vendor's ownership changes?

Cited in the file, not printed as codes: R04, R07, R13

Do not put a failure rate in a business case. No independent study measures how often a PIM project fails. Any round number of that kind in the 2025 edition was cut.

The nearest independent look is a 2021 study by ZHAW of more than 100 companies in DACH. It is stale. Respondents named insufficient staff at about 10%, lack of PIM expertise at about 9%, and then unclear responsibilities, a complex implementation and unwillingness to change processes at about 7–8% each (ZHAW 2021). Read it as a list of causes from that sample, not as a rate you can import.

Our own whitepaper names seven mistakes: no strategy, a process the software does not fit, the wrong stakeholders, data that is not ready, integration left too late, a scope that moves, and training that never happens. Those are our experience. They are not statistics. They line up with the ZHAW list on people, process and ownership, which is why we still use them.

A vendor will show you a feature gap. The stalls we can evidence are earlier. Nobody owns the record. The people who know the data are not on the project. The process stays the spreadsheet, with a new interface in front of it.

Cited in the file, not printed as codes: R13 S44, R13 S03

Completeness is the share of required attributes that are filled, for a channel you name. An attribute that a marketplace does not require is not part of that share.

Validity is whether a value is a legal value. A GTIN with a wrong check digit is invalid even when the field is full.

Consistency is whether the same fact matches across channels. The material on the shop and the material in the marketplace feed are one test.

Timeliness is how old the price and the availability are, against a refresh rule you set. Chapter 05 noted that the AI feed specifications do not set that interval for you.

Channel readiness is the share of SKUs that meet the completeness, validity and image rules of one channel. One vendor index uses "at least 85% publish-ready" as its own benchmark (inriver 2025). That is a vendor claim about its index. It is not a target for your programme. No independent benchmark of these five was found.

ISO 8000 is the standard family the 2025 edition pointed at for data quality. The parts we could identify are ISO 8000-1:2022 and ISO 8000-100:2016 (ISO 2022; ISO 2016). We do not summarise them. The text is paywalled, and earlier sentences about lineage and profiling could not be checked, so they are gone.

Set the five definitions with your team before the vendor's project plan is signed. A measure invented after go-live will describe the project you had, not the one you needed.

Cited in the file, not printed as codes: R09 S4, R14 S53, R05 S46

1. How to read a profileWhat the tiles, the dots and the case page mean.
2. The profilesOne page each, in alphabetical order. A case page only where a case passed the check.

A vendor is in this chapter for one of three reasons. It was in the 2025 edition, which accounts for 28 profiles. Or our coverage scan scored it 6 or higher out of 8, which added crossbase, Struct, Goaland and Quable. Or it was added because European buyers were already meeting it in a 2026 evaluation: Salsify, Plytix, apollon, Pimberly, Agility PIM, Informatica Product 360, Equadis and Base.com. Base.com is still an open decision, because its centre of gravity is syndication. We do not print placements from gated research. Being in this list is not a score.

Names have changed since the 2025 edition. Contentserv is now Centric PXM. Dynamicweb is now Truvio. Informatica Product 360 belongs to Salesforce. Perfion belongs to Boyum IT. SRC belongs to AGENA3000, dated 15 May 2024.

A profile page has the same parts for every vendor. Four tiles: founded, headquarters, deployment, and how many of twelve AI capabilities are documented as generally available. Three columns: what the product is, where it fits, and which modules the vendor names. Dots for AI status. Up to four customer logos, each a vendor claim. A "Good to know" box with one change, one fact and one question to ask. A second page only if a customer case passed the rules. Results on that page are vendor claims.

Order is alphabetical. No vendor reviewed its own profile before this draft. Where two sources disagreed, the page keeps the higher one and the conflict sits in the production notes.

Cited in the file, not printed as codes: R03, R02, R07

Founded in 2013, Akeneo Product Cloud is the product Akeneo sells for managing and publishing product information. The company is based in France. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, syndication. Acquired PricingHUB (announced 2026-06-11); Stripe agentic-commerce partnership (2026-02-18); Summer Release 2026 'Agentic Ziggy' (2026-07-08); Product Cloud adds SDM; PX Insights; Digital Showroom; RCS Nantes 790

Product: Akeneo Product Cloud. Licence: Open Source. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: rewrite. Source date: 2020.

Scope

Direct-to-consumer and wholesale product content.

Challenge

Babolat wanted one product record for its channels, with R&D, product and sales on the same model.

What was done

Akeneo PIM. The vendor's page still states an error-rate figure that cannot be a reduction, so the figure is omitted.

Results

No result figures are printed on this page.

Founded in 2017 in report; company AtroCore GmbH 'founded 2020' per vendor [S45]: product vs company year, needs vendor confirmation, AtroPIM is the product AtroPIM sells for managing and publishing product information. The company is based in Germany. It names manufacturers, wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM, syndication, feed management. Company address on legal notice differs from report; AtroPIM.

Product: AtroPIM. Licence: Open Source and Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Founded in 2016 in report; OmniCX Digital incorporated 3 Nov 2015 (as Saambukapu Ltd, renamed 17 Nov 2016) [S48]; third-party summary says 2015: needs vendor confirmation, BetterCommerce is the product BetterCommerce sells for managing and publishing product information. The company is based in United Kingdom. It names retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, DXP/CMS, commerce. Positioning: B2B commerce platform with.

Product: BetterCommerce. Licence: n/a (the licence row reads 'API-first and Cloud Native Software'). Audience named: retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Comprehensive omnichannel solution unifying systems, eliminating silos and modernizing operations

Challenge

Previous Magento-based platform with limited functionality, fragmented tech stack and lack of modern features for engagement and management

What was done

Comprehensive omnichannel solution unifying systems, eliminating silos and modernizing operations

Results

• vendor claim

Founded in 2016 (vendor [S38]); legal entity registered 2000, renamed 2018 [S36], Bluestone PIM is the product Bluestone PIM sells for managing and publishing product information. The company is based in a headquarters this pass could not confirm. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM. AI Agent launched 2025-10-21; Starboard case updated 2025-04-10; Digital Product Passport release.

Product: Bluestone PIM. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Bluestone PIM

Challenge

Starboard's products are very seasonal and require constant updates; needed flexibility and ease of collaboration between teams and scalability to grow globally

What was done

Bluestone PIM

Results

• vendor claim

Founded in 2001 in report; Handelsregister data shows GmbH founded 2002-01-04 [S57]; Noerr 2017 says AG founded 2001 [S53]: legal-form change, needs vendor confirmation, censhare Digital Experience Platform is the product censhare sells for managing and publishing product information. The company is based in Germany. It names manufacturers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, syndication, DXP/CMS. Positioned inside Entirely ecosystem.

Product: censhare Digital Experience Platform. Licence: Proprietary Software. Audience named: manufacturing, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Holistic platform addressing agile data routing, content management and DAM, integrated with PIM; requirements aligned via stakeholder workshops

Challenge

~7,000 B2B customers, 5+ online shops and 500,000+ product data assets; needed to ensure quality, relevance and cost-effective operations in a fast-changing digital market

What was done

Holistic platform addressing agile data routing, content management and DAM, integrated with PIM; requirements aligned via stakeholder workshops

Results

• vendor claim

Founded in Contentserv founded in 2000 per Dassault and McDermott [S18,S19]; report printed 1999 (needs correction or vendor explanation), Centric PXM is the product Centric PXM (Centric Software) sells for managing and publishing product information. The company is based in United States. It names manufacturers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, syndication, feed management. Acquisition announced 2025-02-25; rebranded Centric PXM.

Product: Centric PXM. Licence: Proprietary Software. Audience named: manufacturing, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Centric PXM for product information and DAM, Translation Manager, Workflow Management

Challenge

Needed to centralize, manage and automate product information and digital assets, with translation and workflow management

What was done

Centric PXM for product information and DAM, Translation Manager, Workflow Management

Results

• vendor claim

Founded in 2008 in report; ID interactive renamed ConnectingTheDots Oct 2017 and was 'ten years old' (implies ~2007, inference) [S62]: needs vendor confirmation, ConnectingTheDots PIM is the product ConnectingTheDots sells for managing and publishing product information. The company is based in Netherlands. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, syndication, feed management. Qmica partnership announced 2026-01-27; Emerce 100.

Product: ConnectingTheDots PIM. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Founded in 1999 [S4], Truvio (formerly Dynamicweb) is the product Truvio (formerly Dynamicweb) sells for managing and publishing product information. The company is based in Denmark. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, DXP/CMS, commerce. Brand being replaced by Truvio (launch 21 Jan 2026); product marketed as Truvio Commerce / Truvio PIM; dynamicweb.com still live

Product: Truvio (formerly Dynamicweb). Licence: NOT FOUND on current site; needs vendor confirmation. Audience named: manufacturing, wholesale distribution, retail brands.

Founded in 2018 (print); legal entity registered 2021 [S73]; needs vendor confirmation, Ergonode is the product Ergonode sells for managing and publishing product information. The company is based in Poland. It names retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, syndication. Deployment in print ("On Premise") conflicts with vendor case page and site (SaaS/cloud); free tier and tiered pricing now published; AI Photo.

Product: Ergonode. Licence: Open Source (print). Free and paid plans, Enterprise on request [S13]; GitHub repos OSL-3.0 but dormant since Feb 2023 [S16]; needs vendor confirmation. Audience named: retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Ergonode PIM: cloud-based, multilingual solution with workflow and user management

Challenge

Decentralized product data, low quality and siloed processes; needed to centralize information, improve multilingual content and enhance team collaboration

What was done

Ergonode PIM: cloud-based, multilingual solution with workflow and user management

Results

• vendor claim

Founded in 2005 = development company start (Bintime); Gepard platform from 2015, SaaS multi-tenant 2021 [S21], Gepard is the product Gepard sells for managing and publishing product information. The company is based in Netherlands. It names wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, syndication. Repositioned "AI-first

Product: Gepard. Licence: Proprietary (print); freemium agreement on site [S19]; needs vendor confirmation. Audience named: wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Gepard PIM integrating ERP system and EPREL; real-time synchronization, standardized product formats, regulatory compliance, automated processes

Challenge

Enhance time-to-market and ensure accurate product information delivery to the EPREL database

What was done

Gepard PIM integrating ERP system and EPREL; real-time synchronization, standardized product formats, regulatory compliance, automated processes

Results

• vendor claim

Founded in 2007 [S26], Inriver PIM is the product inriver sells for managing and publishing product information. The company is based in Sweden. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, syndication. Main phone differs from print; agentic AI release Mar 2026; Microsoft Certified Software May 2026 and SAP S/4HANA connector certified May 2026; Summer 2026 release

Product: Inriver PIM. Licence: Proprietary (SaaS). Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Inriver PIM platform covering SKU creation and enrichment to publication

Challenge

Modernize operations, replacing outdated systems; centralize all product information (19 web stores in 13 languages, 14 retail stores in six countries, 280,000+ SKUs)

What was done

Inriver PIM platform covering SKU creation and enrichment to publication

Results

• vendor claim

Founded in 2017 (product) [S38]; company B.V. incorporated 2003 [S76], KatanaPIM is the product Katana PIM sells for managing and publishing product information. The company is based in Netherlands. It names manufacturers, wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, syndication. No change in contact data; release train 2.12 (Nov 2025) to 2.21 (Apr 2026); case pages OOSC and Polewolf last modified.

Product: KatanaPIM. Licence: Proprietary (print); monthly/yearly subscription plans [S43]. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

OOSC: KatanaPIM as central platform for product data, translations, images and specifications; ChannelEngine integration for marketplace sync. || Polewolf: scalable PIM with fast onboarding and full catalog migration, training and support, Shopify and marketplace integration, validation rules and workflows

Challenge

OOSC: managing expanding product content across multiple online marketplaces; manual handling resulted in errors and inefficiencies. || Polewolf: fragmented product information in spreadsheets led to inconsistent descriptions and slow time-to-market

What was done

OOSC: KatanaPIM as central platform for product data, translations, images and specifications; ChannelEngine integration for marketplace sync. || Polewolf: scalable PIM with fast onboarding and full catalog migration, training and support, Shopify and marketplace integration, validation rules and workflows

Results

• vendor claim

Founded in 1996 per company blog [S79]; register entity from 2005 [S67]; needs vendor confirmation, mediacockpit is the product mediacockpit sells for managing and publishing product information. The company is based in Germany. It names an audience this pass could not confirm. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, syndication. Release 25.3 with expanded AI functions; Alanko AutomotivePIM story published Apr 2025; Linde and.

Product: mediacockpit. Licence: Proprietary (print). Audience named: NOT FOUND.

Case status in the file: use. Source date: 2025 edition.

Scope

Bertsch Innovation's AutomotivePIM as central product information system; automated import, maintenance, enrichment and output; automatic link transfer between turbocharger and engine to vehicles, restrictions on vehicles

Challenge

Maintenance and management of all product data, media content and relationships between vehicles and parts

What was done

Bertsch Innovation's AutomotivePIM as central product information system; automated import, maintenance, enrichment and output; automatic link transfer between turbocharger and engine to vehicles, restrictions on vehicles

Results

• vendor claim

Founded in 2000 [S69,S70], novomind iPIM is the product novomind iPIM sells for managing and publishing product information. The company is based in Germany. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, syndication, commerce. Named in Forrester PIM Systems Landscape Q3 2026 (6 Aug 2026); new customers announced: LG Electronics Deutschland (11 Aug 2026); EFB-Elektronik (17 Sep 2026); AI

Product: novomind iPIM. Licence: Proprietary (print). Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

novomind iPIM with hundreds of automated clean-up and validation rules and a double-checking system

Challenge

Handling 1.3 million article data sets from 250+ suppliers across 1,000+ catalogs; needed a streamlined onboarding process

What was done

novomind iPIM with hundreds of automated clean-up and validation rules and a double-checking system

Results

• vendor claim

Founded in 1998, OmniStock is the product OmniStock sells for managing and publishing product information. The company is based in Netherlands. It names wholesalers, retailers and brands. The stated segment is Midmarket. Native modules confirmed here: PIM. Site presents OmniStock primarily as omnichannel retail logistics (OMS + PIM + smart stock); PIM is one of three solutions

Product: OmniStock. Licence: Proprietary Software. Audience named: wholesale distribution, retail brands.

Founded in 1997, Perfion is the product Perfion sells for managing and publishing product information. The company is based in Denmark. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, syndication. Website moved to perfion.boyum-solutions.com (2023-05-05)

Product: Perfion. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Perfion PIM as central product data repository; Print Publishing module; ERP (Lawson M3) and CMS (Umbraco) integration enabling an online product configurator

Challenge

Data fragmentation and inefficiencies in product data handling

What was done

Perfion PIM as central product data repository; Print Publishing module; ERP (Lawson M3) and CMS (Umbraco) integration enabling an online product configurator

Results

• vendor claim

Founded in 2015, Pimalion is the product Pimalion sells for managing and publishing product information. The company is based in Belgium. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, MDM, syndication. Street address differs between report and current contact page (2015-07-02)

Product: Pimalion. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

Pimalion: bulk imports, Excel-based batch operations, API-driven distribution to eCommerce, POS, Google Shopping and internal tools

Challenge

Optimize product data management: automate tasks, centralize content, support faster decision-making

What was done

Pimalion: bulk imports, Excel-based batch operations, API-driven distribution to eCommerce, POS, Google Shopping and internal tools

Results

• vendor claim

Founded in 2009, Pimcore is the product Pimcore sells for managing and publishing product information. The company is based in Austria. It names manufacturers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, MDM, syndication. Ownership: Tenzing deal (2026-09-02)

Product: Pimcore. Licence: Pimcore Open Core License (POCL), Community Edition free; was printed as "Open Source Software" [S11]. Audience named: manufacturing, retail brands.

Founded in 2019, ProductFlow is the product ProductFlow sells for managing and publishing product information. The company is based in Netherlands. It names wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, syndication. Phone differs from report (+31 85 3012 283)

Product: ProductFlow. Licence: Proprietary Software. Audience named: wholesale distribution, retail brands.

Founded in 2013, Sales Layer is the product Sales Layer sells for managing and publishing product information. The company is based in Spain. It names manufacturers, wholesalers, retailers and brands. The stated segment is Midmarket. Native modules confirmed here: PIM, DAM, syndication. Street address now available (report: city only) (2025-12-11)

Product: Sales Layer. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: rewrite. Source date: NOT FOUND.

Scope

Online sales.

Challenge

The result is a customer quotation on the vendor's homepage, not an independent measurement.

What was done

Sales Layer. Attribute the words to María Selma, eCommerce manager at Ysabel Mora, via the vendor page.

Results

• vendor claim
• vendor claim

Founded in 2004, Signifikant is the product Signifikant sells for managing and publishing product information. The company is based in Sweden. It names manufacturers. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, syndication, commerce. Address; phone; email unchanged (2004-03-12). Figures on this page are sourced in the profile record and are not a ranking.

Product: Signifikant. Licence: Proprietary Software. Audience named: manufacturing.

Case status in the file: use. Source date: 2025 edition.

Scope

Signifikant Aftermarket Platform as single source of truth for spare parts catalogs, exploded views, technical documentation and compatibility data; integrated with ERP and PLM

Challenge

Accurate and up-to-date service information across an international dealer and service network; thousands of parts and extensive technical documentation

What was done

Signifikant Aftermarket Platform as single source of truth for spare parts catalogs, exploded views, technical documentation and compatibility data; integrated with ERP and PLM

Results

• vendor claim

Founded in 2017, Skwirrel is the product Skwirrel sells for managing and publishing product information. The company is based in Netherlands. It names wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM, syndication. No change found. Figures on this page are sourced in the profile record and are not a ranking.

Product: Skwirrel. Licence: Proprietary Software. Audience named: wholesale distribution, retail brands.

Founded in 1987 (report); site says "30+ years" [S3], SRC is the product SRC sells for managing and publishing product information. The company is based in Netherlands. It names manufacturers, wholesalers, retailers and brands. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM. Acquisition by AGENA3000 announced 2024-05-15; before the 2025 edition was made; the 2025 profile did not mention it

Product: SRC. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: use. Source date: 2025 edition.

Scope

SRC plug-and-play supplier portal with ETIM-based classification

Challenge

Managing vast volumes of product data; limitations of spreadsheet-based systems

What was done

SRC plug-and-play supplier portal with ETIM-based classification

Results

• vendor claim

Founded in 1976, Stibo Systems Multidomain MDM is the product Stibo Systems sells for managing and publishing product information. The company is based in Denmark. It names manufacturers, wholesalers, retailers and brands. The stated segment is Enterprise. Native modules confirmed here: PIM, DAM, MDM, syndication. MCP Server (2026-06-29); AgentWorkx (2026-09-29); record fiscal year (2026-06-25); Bosch partnership (2026-06-11); Bygma (2026-09-15); 50-year milestone (2026-05-05; consistent with 1976 founding); no acquisitions or ownership

Product: Stibo Systems Multidomain MDM. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution, retail brands.

Case status in the file: rewrite. Source date: NOT FOUND.

Scope

Product data for retail expansion.

Challenge

Use only figures on Stibo's own Rituals page. The 2025 edition copied another vendor's results.

What was done

Stibo Systems. The 353% and 71% figures are Sales Layer's and are not printed here. The 1000% store-openings claim is omitted.

Results

• vendor claim
• vendor claim
• vendor claim
• vendor claim

Founded in 2003 (report); SyncForce site traces origin to 1999 (founder Hans de Gier; Deloitte Fast 50 in 2001) [S27]: CONFLICT, needs vendor confirmation, SyncForce is the product SyncForce sells for managing and publishing product information. The company is based in Netherlands. It names manufacturers. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM, syndication, PLM. Address/phone unchanged; plus Strijp-T; TQ5; 5th floor; moved to Strijp-T.

Product: SyncForce. Licence: Proprietary Software. Audience named: manufacturing.

Case status in the file: use. Source date: 2025 edition.

Scope

SyncForce 'Circular PIM' across product and packaging domains: real-time label updates, compliance modules, GS1 alignment; started with workshops

Challenge

Data quality under manual processes and threat of non-compliance with GS1 standards; retailers such as Albert Heijn signalled penalties or delisting

What was done

SyncForce 'Circular PIM' across product and packaging domains: real-time label updates, compliance modules, GS1 alignment; started with workshops

Results

• vendor claim

Founded in 1987, Syndigo is the product Syndigo sells for managing and publishing product information. The company is based in United States. It names manufacturers, retailers and brands. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, MDM, syndication. Acquired 1WorldSync; announced 2025-09-03; combined EV >$3.5B per Summit Partners page; 1; 500+ employees; 12 offices; 18; 000+ customers in 60 countries (claims)

Product: Syndigo. Licence: Proprietary Software. Audience named: manufacturing, retail brands.

Founded in 1984, Utopis PIM is the product Utopis sells for managing and publishing product information. The company is based in Netherlands [S10]. It names manufacturers, wholesalers. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM. No change in address/phone/email; re-confirmed 2026-10-08

Product: Utopis PIM. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution.

Founded in 1999, Viamedici EPIM is the product Viamedici sells for managing and publishing product information. The company is based in Germany. It names manufacturers, wholesalers. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM, MDM, syndication. published 2026-04-06; MDM not PIM; 300+ market leaders; 99% retention are vendor claims; Bloor 2024 gave >400 customers; 200 staff

Product: Viamedici EPIM. Licence: Proprietary Software. Audience named: manufacturing, wholesale distribution.

Founded in 2012, Salsify is the product Salsify sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, syndication. Cinven announced Salsify on 2026-07-22

Product: Salsify. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2015, Plytix is the product Plytix sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM, feed management. No dated change since the 2025 edition was confirmed in this pass

Product: Plytix. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2015, apollon OMN is the product apollon OMN sells for managing and publishing product information. The company is based in in DACH. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, DAM, syndication. No dated change since the 2025 edition was confirmed in this pass

Product: apollon OMN. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2015, Pimberly is the product Pimberly sells for managing and publishing product information. The company is based in United Kingdom. It names an audience this pass could not confirm. The stated segment is Midmarket / Enterprise. Native modules confirmed here: PIM, DAM. Virgin Money (lender) announced Pimberly on 2025-12-17

Product: Pimberly. Licence: Proprietary Software. Audience named: NOT FOUND.

Founded in 2010, Agility PIM (insightsoftware) is the product Agility PIM (insightsoftware) sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Agility PIM (insightsoftware). Licence: Proprietary. Audience named: NOT FOUND.

Founded in 1987, Informatica Product 360 (Salesforce) is the product Informatica Product 360 (Salesforce) sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Enterprise. Native modules confirmed here: PIM, MDM, syndication. Salesforce announced Informatica (incl. Product 360 PIM and MDM) on 2025-05-27

Product: Informatica Product 360 (Salesforce). Licence: Proprietary. Audience named: NOT FOUND.

Equadis is the product Equadis sells for managing and publishing product information. The company is based in a headquarters this pass could not confirm. It names an audience this pass could not confirm. The segment is not confirmed in this pass. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Equadis. Licence: NOT FOUND. Audience named: NOT FOUND.

Base.com is the product Base.com sells for managing and publishing product information. The company is based in a headquarters this pass could not confirm. It names an audience this pass could not confirm. The segment is not confirmed in this pass. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Base.com. Licence: NOT FOUND. Audience named: NOT FOUND.

Founded in 2001, crossbase is the product crossbase sells for managing and publishing product information. The company is based in < 5 countries. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: PIM, syndication, DXP/CMS. No dated change since the 2025 edition was confirmed in this pass

Product: crossbase. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2015, Struct PIM is the product Struct PIM sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Struct PIM. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2016, Goaland is the product Goaland sells for managing and publishing product information. The company is based in in Europe. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Goaland. Licence: Proprietary. Audience named: NOT FOUND.

Founded in 2013, Quable is the product Quable sells for managing and publishing product information. The company is based in globally. It names an audience this pass could not confirm. The stated segment is Entry-level / Midmarket. Native modules confirmed here: no native module list confirmed. No dated change since the 2025 edition was confirmed in this pass

Product: Quable. Licence: Proprietary. Audience named: NOT FOUND.

Research for this edition ran from 8 October to 17 October 2026, in seventeen tasks. Pages that were checked again during production carry their own access date. The facts were closed on 17 October 2026 unless a page says otherwise.

We preferred primary sources: the Official Journal, standards bodies, and the vendor's own documentation or press release. A vendor's number is printed as a vendor claim. A figure we could not find is left as a gap. An inference is marked as our reading. Confidence is high when the source is primary and recent, medium when it is a solid secondary source or an older primary, and low when it rests on one weak source.

We did not run a buyer survey. Gaps on cost, maturity and why projects fail stay labelled as gaps. Gated analyst reports are cited only where a public page confirms a name. Placements from gated research are not printed. No vendor paid for a profile in this draft, and no vendor reviewed its page before it.

We checked 47 claims in the 2025 edition. Eight were verified. Thirteen were out of date. Seven did not match their source. One was false. Eighteen could not be traced. The reference list at the back is rebuilt from the sources this edition actually uses.

Chapters that were not in the 2025 edition are the AI maturity ladder, agentic commerce, the regulation timeline, cost and implementation, Benelux and DACH, the buyer checklist, the glossary, and twelve of the forty profiles.

Cited in the file, not printed as codes: R05

PIMvendors.com is an independent platform for product information management. We fund the work through success fees, so the advice stays independent of a licence we do not sell. Inclusion in this report is not an endorsement. We do not rank vendors.

Chris Jobse is founder and co-owner. He advises companies on master data and product information, on the process around the data, and on the tooling that has to support it. In 2022 he was appointed Shopping Award Vakexpert.

Stephan Spijkers is founder and co-owner. He has more than ten years in online retail, selling digital products, subscriptions and physical goods. As chief operating officer of Westwing Benelux he worked with a team of about 50 people whose job was to get product data online, launching more than 1,200 new products a day.

The report is published by PIMvendors.com B.V. Questions on this edition: report@pimvendors.com.