Amazon is considering building a database of every product sold on the internet, including products it does not sell, according to Reuters, which reported on an internal company document.
The project, called the "World Product Graph," is intended to reduce the likelihood that shoppers leave Amazon to research products on platforms like Reddit or Google. If Amazon has the largest stockpile of shopping data, AI agents may rely on its product index over others, so people shopping with bots could end up on its pages.
Links to outside retailers' products have already begun appearing on Amazon through a "shop direct" feature launched last year. The company also operates its own AI shopping assistant, Alexa for Shopping, formerly called Rufus, which the World Product Graph would aim to improve.
The internal document reportedly acknowledges shortcomings in the current system, stating that "Many Rufus answers are factually wrong or incomplete" and that running users' queries through the AI is expensive. Adding more data is reportedly supposed to make the system more accurate and able to use cheaper models.
Reuters reported that the document included a release date of October 1, 2027, but it remains unclear if or when the plan may be implemented. The project reportedly depends on discovering how brands organise their products on their own websites, which could involve Amazon crawling other retailers' sites.
In response to Reuters, Amazon said: "Draft, outdated documents often paint an incomplete picture and the existence of a brainstorming document such as this one doesn't necessarily reflect our current plans."
Analysis
Why This Matters
- If implemented, Amazon's index could reshape online shopping by consolidating product discovery on its platform, potentially reducing traffic to competing retailers and review sites.
- The project highlights Amazon's ambition to dominate the AI shopping assistant market by controlling the underlying product data, similar to how search engines rely on web indexes.
- The plan's dependence on crawling competitors' sites raises questions about data ownership, fair use, and possible antitrust scrutiny.
Background
Amazon already operates a "shop direct" feature that links to outside retailers, and its AI shopping assistant, Alexa for Shopping (formerly Rufus), provides product answers. The World Product Graph would expand on these efforts by creating a unified index covering products not sold on Amazon, aiming to solve accuracy and cost issues with the current AI system.
Key Perspectives
Amazon: says the document is an incomplete draft and does not reflect current plans.
Internal Amazon document: states that the system is needed because current Rufus answers are often factually wrong or incomplete, and that the project could make Amazon the default source for product data used by AI agents.
What to Watch
- Whether Amazon publicly confirms further development of the World Product Graph or announces a pilot program.
- Reactions from retailers, regulators, and competitors (such as Google Shopping or Shopify) to the prospect of Amazon crawling their product pages.
- Updates to Amazon's AI shopping assistant that might indicate the project is moving forward, such as improved accuracy or lower operational costs.
Checked
12 claims in this story checked against 1 source: 12 supported.
Show each claim
- SupportedAmazon has a plan to index products sold across the internet, including products it does not sell, according to Reuters.The Verge: “Amazon has plans to build a vault of data on every product sold anywhere on the internet, even products Amazon doesn’t sell, and make the information available through its own site, according to Reuters.”
- SupportedThe project is called the “World Product Graph.”The Verge: “The project, called the “World Product Graph,””
- SupportedThe project is intended to make shoppers less likely to leave Amazon to research products on Reddit or Google.The Verge: “is meant to make it less likely that shoppers leave Amazon to do their product research on Reddit or Google”
- SupportedIf Amazon has the largest stockpile of shopping data, AI agents may rely on its product index over others, potentially directing shoppers using bots to Amazon’s pages.The Verge: “if Amazon has the largest stockpile of shopping data, AI agents may rely on its index of products over others, so even people shopping with bots will end up on its pages.”
- SupportedLinks to outside retailers’ products have started appearing on Amazon through its “shop direct” feature, launched last year.The Verge: “Links to outside retailers’ products have already started popping up on Amazon through the “shop direct” feature it launched last year.”
- SupportedAmazon’s AI shopping assistant is called Alexa for Shopping and was previously called Rufus.The Verge: “The company also has its own AI shopping assistant, Alexa for Shopping, previously called Rufus.”
- SupportedAn internal Amazon document says, “Many Rufus answers are factually wrong or incomplete”.The Verge: “According to Reuters, an internal Amazon document on the project says “Many Rufus answers are factually wrong or incomplete””
- SupportedThe document says running users’ Amazon searches with the AI is expensive.The Verge: “and running users’ Amazon searches with the AI is expensive.”
- SupportedAdding more data is reportedly intended to make the system more accurate and able to use cheaper models.The Verge: “Adding more data is reportedly supposed to make the system more accurate and able to use cheaper models.”
- SupportedThe document included an October 1, 2027 release date, but it is unclear if or when the plan may be implemented.The Verge: “Reuters says it’s unclear “if or when this plan may be implemented,” but said the document on it included an October 1st, 2027 release date.”
- SupportedThe project reportedly depends on discovering how brands organise products on their own websites, which could involve Amazon crawling other retailers’ sites.The Verge: “The internal document reportedly says the project “depends on discovering how brands organize their products on their own websites,” which could involve Amazon crawling other retailers’ sites.”
- SupportedAmazon told Reuters that draft, outdated documents can be incomplete and that the brainstorming document does not necessarily reflect current plans.The Verge: “Amazon said in a statement to Reuters 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.””
