Amazon has considered a plan to index online shopping data across the internet through a project called the “World Product Graph,” according to internal documents reviewed by Reuters on Friday.
The initiative aims to collect information on products the company does not sell itself, helping consumers compare items and allowing AI shopping assistants to retrieve deeper product relationships.
While Amazon commands a strong position when shoppers know precisely what they want to buy, early-stage product discovery presents a distinct vulnerability. Industry data cited in recent reports shows that roughly 57 percent of shoppers begin their open-ended product research on Google, brand websites, or review forums, compared to only 16 percent on Amazon.
By contrast, when customers have already decided on a specific purchase, Amazon captures roughly 41 percent of searches, outpacing Google at 23 percent and Walmart at 18 percent.
Amazon cannot be the go-destination for every shopping mission if customers must leave to research product relationships on Google, brand websites, and review sites.
Internal Amazon document, via Reuters
Company documents indicate that customers have always been able to find products on Amazon, but that understanding how products relate to one another has been missing from the experience. The World Product Graph initiative seeks to address this gap by gathering data on virtually every item for sale on Earth, cutting into Google’s dominance in online search while fending off growing e-commerce competition from Walmart.
The push for a comprehensive product index also ties directly into the performance of Amazon’s AI shopping assistant, Alexa for Shopping, which was previously called Rufus. Internal documents said that many Rufus answers are factually wrong or incomplete, and running user searches through the AI remains expensive.
Enriching product data is reportedly intended to make the system more accurate and allow it to utilize cheaper AI models efficiently. The project is designed to ensure that if AI agents take on a larger role in autonomous shopping, those bots will rely on Amazon’s massive stockpile of product data rather than competing indexes, keeping users on Amazon pages even when searching for items sold elsewhere.
The proposed indexing effort builds upon existing services such as Shop Direct and Buy for Me, which already list outside products on Amazon and allow customers to complete purchases using Amazon as an intermediary. Shop Direct, which began in 2025, covered more than 100 million products from over 400,000 outside stores as of March 11, 2026, according to company figures.
The internal planning document listed a desired launch date of October 1, 2027, for the World Product Graph. However, the project depends on discovering how brands organize their products on their own websites, which could involve automated crawling of third-party store pages alongside data feeds provided directly by brands.
In response to the Reuters report, Amazon said 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.
