Amazon has a plan to index the world’s online shopping data
The Verge
Amazon is reportedly developing the World Product Graph to index global online shopping data and keep users on its platform.
According to a Reuters report, Amazon is developing a project known as the World Product Graph to build a comprehensive database of every product sold across the entire internet, including items not sold directly by Amazon. The initiative aims to prevent shoppers from leaving Amazon to conduct research on platforms like Google or Reddit, while also ensuring that emerging AI shopping agents rely on Amazon's data index.
The project is also intended to address internal challenges with Amazon's AI shopping assistant, previously called Rufus and currently known as Alexa for Shopping. Internal company documents reportedly noted that many responses generated by the assistant are incomplete or factually incorrect, and that utilizing the AI is costly. By integrating a broader index of product data, Amazon hopes to improve the assistant's accuracy while enabling the use of cheaper models.
While the internal documents pointed to an October 1, 2027 release date, implementation remains uncertain. The feasibility of the project depends on Amazon's ability to discover how various brands organize products on their own websites, which could require crawling external retailer sites. In response to the report, Amazon stated that outdated draft documents do not necessarily reflect current company plans.
Why it matters: The initiative could significantly expand Amazon's reach over online commerce and AI-driven shopping, ensuring that consumers using bots or conducting product research are funneled back to its platform.

