Skip to main content

Amazon is developing a proposed World Product Graph

14 NOVEMBER 2026·2 MIN READ·2 SOURCES

Amazon is developing a proposed World Product Graph that would collect information on products worldwide, including items it does not sell, Reuters reported on Nov. 13. The plan would depend on data from brands and retailers as well as Amazon crawling rival websites to fill gaps.

Amazon is developing a proposed World Product Graph

Key takeaways · 4

  • 01

    The proposed graph would include products Amazon does not sell, with brands and retailers contributing data and Amazon crawling rival sites to fill gaps.

  • 02

    Coca-Cola and Bayer have signed up to contribute data; Procter & Gamble attended a trial summit related to the project.

  • 03

    An internal document listed Oct. 1, 2027, as an aspirational release target, not a confirmed launch date.

  • 04

    Amazon says Alexa for Shopping is roughly 97% accurate globally; the reported weekly search figure of about 83 million is described as significantly outdated.

A broader catalogue for shopping AI

The proposed World Product Graph would gather information about products around the world, including products that Amazon does not sell.[1] The reported strategic possibility goes further: if Amazon holds the largest stockpile of shopping data, AI agents may rely on its product index rather than competing indexes.[2] That is a potential consequence described in the reporting, not evidence that AI agents have already adopted the graph. The project would also provide AI with more data to help address issues with its answers.[2]

Data contributions and open questions

The plan would rely on brands and retailers contributing product information, while Amazon would crawl rival websites to fill gaps.[1] Coca-Cola and Bayer have signed up to contribute data, and Procter & Gamble attended a trial summit related to the project.[1] Amazon could still change or abandon the project, Reuters reported.[1] Amazon also said the internal document describing the plan did not necessarily reflect its current plans.[1]

Rufus figures need careful reading

Amazon has renamed Rufus Alexa for Shopping, and Amazon says the shopping assistant is roughly 97% accurate globally.[1] Separately, an internal document put weekly searches at about 83 million, but that usage figure is described as significantly outdated.[1] The evidence also says many Rufus answers are factually wrong or incomplete.[2] The figures therefore should not be treated as a current measurement of usage or as proof that the assistant’s answers are uniformly reliable. The internal document said richer product data could let Amazon use less expensive models to answer Rufus queries.[1]

A proposed timeline, not a launch commitment

The internal document gave the project an aspirational release target of Oct. 1, 2027.[1] That date is not a confirmed launch commitment: Amazon said the document did not necessarily reflect its current plans, and Reuters reported that the company could modify or abandon the project.[1] Amazon launched Shop Direct last year, allowing links to products from outside retailers to appear on Amazon.[2] That earlier feature offers context for the company’s efforts to surface outside-retailer products, but the evidence does not say Shop Direct is part of the proposed graph.

Retailers and brands considering participation should distinguish a proposed data-sharing effort from a confirmed product launch. For teams evaluating shopping assistants, the reported data-quality issues, outdated usage figure and conditional timeline are reasons to validate current performance and plans before making commitments.

Why it matters
For teams

AI training for your team

kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.

How this developed

  1. 14 November 2026

    Amazon is developing a proposed World Product Graph

Sources

AI fluency, one session a day, built for your work.