Opens in a new tab

Amazon wants the product page for every item in the world: who controls your catalogue data

Written by Trilogi – The Digital Agency
14/11/2026
Logistics warehouse with shelves and pallets full of products from different brands

In short

Amazon is studying how to hold the product page for almost every item for sale, including those it does not sell. If someone else builds your product page, someone else decides what customers see and what AI recommends.

Amazon is studying how to hold the product page for practically every item sold in the world, including the ones it does not sell. For any company that manufactures or distributes, the consequence is direct: if someone else builds your product page, someone else decides what the customer sees and what AI recommends. The answer is not to rush to Amazon, but to make your product data the most complete and reliable on the web.

What happened

An internal Amazon document that came to light this week describes a project called World Product Graph: a database with information on almost every item for sale, whether or not Amazon sells it. The document sets an aspirational launch date of 1 October 2027.

The database would be fed in two ways: with data continuously contributed by brands and retailers themselves, and by crawling other merchants’ websites. According to the document, some large consumer brands have already signed up to share information.

The text is clear about the reason: Amazon cannot be the destination for every purchase if customers have to leave for another site to research the product. And it adds a newer concern: that AI shopping assistants will end up sending traffic to the sites with the most complete data.

Amazon is not starting from scratch. It already runs services that show products from other stores and send customers there. One of them grew from 100 million products in March to 250 million in October. Amazon, for its part, says the document is an old draft that does not necessarily reflect its current plans.

The bigger picture: the battle is no longer for the click, it is for the product page

For twenty years, the online retail fight has been about traffic: ranking on Google, bidding on Amazon, attracting visits to your own site. With AI assistants, customers ask less and delegate more: “find me, for example, certified work boots for under 90 euros”. The assistant does not browse like a person; it reads structured data and compares.

In that scenario, whoever has the best description of each product wins: dimensions, materials, compatibility, certifications, stock, up-to-date price. Amazon, Google and OpenAI are moving in the same direction by different routes. The product page stops being copy for the website and becomes the asset that feeds every channel.

Why it matters to your business

Sales. If the assistant finds an incomplete page of yours, or a distributor’s version with outdated data, it can rule you out without anyone on your team noticing. There is no lost visit to measure; you are simply not in the answer.

Brand and price. When a third party crawls and rebuilds your catalogue, it can mix up references, show outdated prices or attribute features your product does not have. That hurts trust and drives returns.

Dependence. Giving your data to a large platform can bring visibility, but it also hands over knowledge of your catalogue. It is worth deciding deliberately, channel by channel, not by inertia.

Internal cost. In many mid-sized companies, product information is scattered across the ERP, spreadsheets, the technical department and the website. Every new channel multiplies manual work and errors.

What to review now

Before thinking about new platforms, leadership should be able to answer these questions:

  • Where is the “official” version of each product, and who is responsible for keeping it complete?
  • Does the data on your own site match what marketplaces and distributors show? An illustrative example: a furniture manufacturer that publishes one set of dimensions on its site and different ones in three stores that resell it.
  • Which channels receive our catalogue automatically today, and which by hand?
  • What does an AI assistant answer when asked about our best-selling products?
  • What data are we willing to share with each large platform, and in exchange for what?

Amazon’s plan may change or never arrive. The underlying trend, however, does not depend on one document: AI buys with data, and product data is becoming a competitive advantage.

How we approach it at TRILOGI

With news like this, we do not start with technology but with a map: which channels sell each product, where its data comes from and where it breaks. From there we define a single source of truth for the catalogue and connect the website, marketplaces and distributors to it, so the product page is the same and complete everywhere.

The next step is to decide, channel by channel, what gets published and in how much detail, and to regularly check how the big search engines and assistants read that catalogue. That is the work we do in marketplace integration: fewer manual tasks, fewer errors and more control over how others present you.

Every Tuesday, the 5 news stories that matter: subscribe to the newsletter.