Skip to content
Bramble

Technology

Amazon plans "World Product Graph" to hold data on every product, document shows

An internal Amazon document reviewed by Reuters describes "World Product Graph", a plan to gather data on nearly every item for sale worldwide and counter AI shopping agents. Amazon says the document is a draft brainstorm that doesn't necessarily reflect its current plans.

2 outlets · 0L · 1C · 1R First reported Account updated
Image: CNA
Image: The Jerusalem Post

1 / 2

The story, neutrally told

Mixed · 2Amazon has an internal plan called World Product Graph that aims to collect data on virtually every item for sale worldwide, whether or not Amazon stocks it, according to an internal document reviewed by Reuters and people familiar with the matter. Mixed · 2The document argues Amazon cannot be the go-to destination for every shopping mission if customers must leave to research product relationships on Google, brand websites and review sites, and the plan is meant to keep shoppers on Amazon. Mixed · 2The document is a PR/FAQ, typically a six-page internal narrative memo, written like a press release with an aspirational release date of October 1, 2027. Reuters could not determine whether or when the plan will be implemented, and people familiar said it could be modified or aborted due to financial or other concerns.

Mixed · 2Amazon said in a statement that "draft, outdated documents often paint an incomplete picture and the existence of a brainstorming document such as this one doesn't necessarily reflect our current plans", adding that it is constantly working on ideas to make its product catalog comprehensive, accurate and helpful. Mixed · 2The data would come partly from brands and retailers contributing product data continuously and partly from Amazon independently crawling other retailers' websites; Reuters could not determine how the crawling would work or whether permission would always be sought. Mixed · 2According to the document, Amazon has already signed up Coca-Cola and Bayer, among others, to contribute data, and held a "CEO-level P&G/Amazon summit" for a trial with Procter & Gamble; the companies did not respond to requests for comment.

Mixed · 2The project sits in Amazon's Selection & Catalog Systems unit and is organised under Robert Tekiela, a vice president in that division, who did not respond to a request for comment. One source said Amazon is shuffling staff internally to get it started. Mixed · 2The plan is also meant to fix problems with Amazon's AI shopping assistant Rufus, renamed "Alexa for Shopping" in May. The document says many Rufus answers are factually wrong or incomplete, and that about 83 million weekly searches are "expensive" to run; more data would allow cheaper models. Amazon said the usage figures are significantly outdated and that Alexa for Shopping is roughly 97% accurate globally. Mixed · 2People familiar said the plan would also counter third-party AI shopping agents that could send traffic to sites with more complete data, and by making Amazon the richest data source would draw such agents to Amazon.com and help protect its advertising business. CEO Andy Jassy told the third-quarter earnings call that "few customers are doing much shopping this way yet".

Mixed · 2Meta's Muse agent has more than 10 million users in the US and Canada since launching in September; Amazon blocked it that month, and last year sued Perplexity over similar access, a case still ongoing. Elon Musk wrote on X, after a post saying xAI's Grok agent was blocked on Amazon: "Maybe we should start a competitor to Amazon." Mixed · 2Context from a 2025 Morgan Stanley report: 16% of online shoppers start product research on Amazon versus 57% on Google, but when shoppers know what they want, 41% of searches start on Amazon, compared with 23% for Google and 18% for Walmart. Mixed · 2Employees describe the project as the start of an effort to make the site a one-stop shop, expanding the existing "Buy for Me" and "Shop Direct" services that list products sold on other sites with Amazon as intermediary.

Every sentence links to the reporting it rests on. The pill in front of each says where its sources sit: Left, Centre or Right when one side supplies at least half of them, Mixed when they are evenly split. The number is how many outlets it cites.

Left0 outlets

No left outlet in our sources has covered this story yet.

Centre1 outlet

Framing
Reuters exclusive carried under CNA's byline, leading with Amazon's "secret plan" to own product data and fight AI shopping agents.
Emphasis
The competitive threat from AI agents, Amazon's own Rufus accuracy problems, and Amazon's rebuttal.
Leaves out or plays down
No reporting beyond the Reuters story; unresolved questions (crawling permissions, timing) are noted as unknown.
Charged language
“secret plan”“hoover up”
For example
“Amazon has a secret plan to ensure shoppers never need to leave its site: own the world’s product information.” — CNA
“Exclusive-Amazon’s plan to own the world’s product data and fight AI shopping agents” — CNA

Right1 outlet

Framing
Reuters copy run by The Jerusalem Post with a headline stressing Amazon amassing data on every item worldwide to fight AI shopping agents, plus added subheads.
Emphasis
Scale of the data collection and the crawling of retailers' sites; subheads on crawling and on pushing AI agents to Amazon's store.
Leaves out or plays down
No added reporting or local angle; the text matches the Reuters wording.
Charged language
“secret plan”“hoover up”“amass data”
For example
“Amazon seeks to amass data on every item for sale worldwide to fight AI shopping agents” — The Jerusalem Post
“Amazon document plans to crawl retailers' websites for product data” — The Jerusalem Post

Ask about this story

Loading…