- What changed
- Internal Amazon PR/FAQ called World Product Graph to catalog nearly every product sold online
- When
- Reported by Reuters November 13, 2026; aspirational release date October 1, 2027
- Who it affects
- Amazon Selection & Catalog Systems; brands such as Coca-Cola and Bayer named as data partners
- What to do
- Audit product data on your own site, Amazon and feeds so every source says the same thing
Key takeaways
- Reuters reported on November 13, 2026 that an internal Amazon plan, World Product Graph, aims to catalog nearly every product sold online by an aspirational October 1, 2027.
- Data would come from brand and retailer feeds plus crawling; Coca-Cola and Bayer are named partners, and Amazon calls the document a brainstorm.
- Make your website, Amazon listings and feeds agree on every key attribute now; it cuts returns and helps AI answers whether or not the plan ships.
Amazon's path to a catalog of everything
Shop Direct starts listing outside stores
Rufus renamed Alexa for Shopping
Amazon blocks Meta's Muse agent
Shop Direct merchant portal opens
Reuters reports World Product Graph
Aspirational release date in memo
What Reuters found
Reuters reporter Greg Bensinger reviewed an internal Amazon document written in the company's PR/FAQ style, a mock press release with questions and answers that Amazon teams use to pitch new products. The plan is called World Product Graph and sits in the Selection & Catalog Systems unit under vice president Robert Tekiela, who did not respond to Reuters. Amazon is moving staff internally to start the work, one source told Reuters.
The goal is a database of nearly every item for sale on Earth. Brands and retailers would send product data continuously, and Amazon would also crawl other retailers' websites on its own. The document says the project depends on learning how brands organize products on their own websites. Reuters could not determine how the crawling would work, whether Amazon would need permission in every case, or whether the October 1, 2027 date is realistic. PR/FAQs are often changed or dropped, and Amazon told Reuters the existence of this one does not necessarily reflect its current plans.
Why Amazon wants it
The document argues Amazon cannot be the place for every shopping mission if customers leave to research products on Google, brand sites and review sites. Reuters cites a 2025 Morgan Stanley report: 57% of shoppers start product research on Google versus 16% on Amazon, though Amazon leads, at 41%, when shoppers already know the exact product they want.
The second reason is AI shopping agents. Reuters reports Amazon executives worry that outside agents will send shoppers to sites with more complete data. Amazon blocked Meta's Muse agent from its store in September and is in an ongoing lawsuit with Perplexity. A fuller catalog would also extend Buy for Me and Shop Direct, the programs that already show products sold on other websites inside Amazon, and help protect Amazon's advertising business, according to people familiar with the plan.
The third reason is Amazon's own assistant. The document says many answers from Rufus, renamed Alexa for Shopping in May, are factually wrong or incomplete and that customers often cannot spot the errors. It puts Rufus at about 83 million searches a week. Amazon told Reuters those usage figures are significantly outdated and that its data shows Alexa for Shopping is roughly 97% accurate globally.
Where shoppers start
We launch and run brands on Walmart, Target Plus, TikTok Shop and more, from one team.
What most coverage misses
Most write-ups frame this as Amazon against Google. For a brand, the sharper point is that Amazon wants to learn your product structure from your own website, not only from your Seller Central listings. Variations, model years, colorways and compatibility notes that exist only on your site, or that disagree with your Amazon detail page, are exactly the gaps the document says Amazon wants to close. If Amazon builds this, whichever source is cleaner and more complete will shape how assistants describe your products, on Amazon and possibly in agents that rely on Amazon data. See our guide to Alexa for Shopping for how the assistant already uses listing content.
So what: the money at stake
Wrong or incomplete product answers cost money twice: lost conversions when a shopper cannot confirm fit, and returns when an assistant gets a spec wrong. The math is simple. Monthly cost of bad data equals units sold times extra return rate times cost per return, plus lost orders.
Worked example, illustrative numbers: a brand sells 2,000 units a month of a $40 blender on Amazon. If an outdated capacity figure on its website conflicts with the Amazon listing and pushes the return rate from 5% to 7%, that is 40 extra returns a month. At about $15 per return in lost fees, shipping and unsellable stock, that is $600 a month, or $7,200 a year, before counting shoppers who never buy. Fixing the data costs one afternoon.
What sellers should do
These steps apply to brand owners selling on Amazon and on their own site. None of them depends on Amazon actually launching the World Product Graph, so they pay off either way.
- Catalog lead: export your Amazon listings and your website product feed, then compare title, size, materials, model year and compatibility for your top 50 ASINs. Metric: number of mismatched attributes, aiming for zero.
- Brand manager: make sure every variation on your site maps to the same parent and child structure on Amazon, using Amazon variations rules. Metric: no orphan children or duplicate listings in Brand Registry.
- Content lead: answer the top 10 questions shoppers ask about each hero product in bullets and A+ content, since assistants pull from that text; our Amazon listing optimization service does this at scale. Metric: fewer Customer Questions and lower return rate for 'not as described'.
- Ecommerce lead: check whether your products already appear through Shop Direct on Amazon and decide whether to manage that feed yourself. Metric: you know every URL Amazon shows for your brand.
- Legal or IT lead: review robots.txt and your site terms so crawling of your product pages is a choice you made, not an accident. Metric: a written policy on which bots may read your catalog.
- Growth lead: track how AI assistants describe your brand once a month, using our guide to brand mentions in AI search. Metric: share of answers with correct specs and price.
Product data checklist
- Compare site feed with Amazon listings
- Match variation families
- Fix model year and size conflicts
- Answer top 10 shopper questions
- Find your Shop Direct URLs
- Set a crawler policy
- Check AI answers monthly
When this does not apply
If you sell only on Amazon and have no website or other retail listings, there is less for Amazon to reconcile, though your listing accuracy still matters for Alexa for Shopping. Resellers and wholesalers cannot control brand data and should focus on matching the brand's official specs. And this is a leaked internal plan, not a launched program: there is no opt-in form, fee or deadline today. Do not change contracts or block crawlers in a panic; make your data consistent, which helps on every channel.
Where PPC Spider helps
Clean product data feeds both organic rank and ad performance, because Sponsored Products relevance depends on the same listing text. Our team runs Amazon account management with catalog checks built in. Start with a free Amazon audit to see where your listings and site disagree, and follow more stories like this on our ecommerce marketplace news channel.
Sources
- Reuters via Investing.com: Exclusive, Amazon's plan to own the world's product data and fight AI shopping agents, November 13, 2026 (https://www.investing.com/news/stock-market-news/exclusiveamazons-plan-to-own-the-worlds-product-data-and-fight-ai-shopping-agents-5024784)
- PYMNTS: Amazon Targets Google and Walmart With Plan to Catalog Every Product Online, November 13, 2026 (https://www.pymnts.com/commerce/ecommerce/2026/amazon-targets-google-walmart-with-plan-catalog-every-product-online/)
- Crypto Briefing: Amazon's World Product Graph plan aims to catalog nearly every product, November 13, 2026 (https://cryptobriefing.com/amazon-world-product-graph-plan-2027/)
- Wilson's Media: Amazon has a plan to index the world's online shopping data, November 13, 2026 (https://www.wilsonsmedia.com/amazon-has-a-plan-to-index-the-worlds-online-shopping-data/)
We run these playbooks on client accounts every day. Start with a free audit.
Get a free audit