Amazon plans World Product Graph project for an index of products from across the entire internet
According to Reuters, Amazon is developing an internal project called World Product Graph – a database of products from across the entire internet, including goods that it does not offer itself. The goal is to keep users on Amazon and improve the AI agent Rufus (now Alexa for Shopping), which, according to an internal document, is often wrong.
According to a Reuters report, Amazon is working on a project called World Product Graph – a database of data on products sold anywhere on the internet, including goods that Amazon itself does not offer. The goal is to reduce the likelihood that users go elsewhere to look for product information on Reddit or Google, and to increase the chance that they end up on Amazon's pages even when searching for goods sold elsewhere. The same is meant to apply to AI agents: if Amazon has the largest supply of shopping data, agents may rely precisely on its index when recommending products.
According to an internal document cited by Reuters, the project is also meant to solve problems with the AI assistant Rufus (now renamed Alexa for Shopping) – the document states that "many of Rufus's answers are factually incorrect or incomplete" and that processing user searches using AI is costly. According to the company, expanding the data base is meant to improve the accuracy of answers and enable the use of cheaper models. Amazon had already previously launched a "shop direct" feature, thanks to which links to products from other sellers appear on its pages.
According to the document, implementation of the project is conditional on Amazon learning how individual brands organize their products on their own websites – which could include crawling the websites of competing sellers. Reuters states that it is not clear whether and when the plan will be implemented, but the document contained a target release date of 1 October 2027.
Amazon told Reuters that "drafts and outdated documents often present an incomplete picture" and that the existence of such a brainstorming document does not necessarily reflect the company's current plans.
Why it matters
For ordinary users, this could mean that AI shopping assistants – whether from Amazon or other companies drawing on similarly extensive indexes – will increasingly steer recommendations toward the platform with the most extensive data base when recommending products, even when the goods being searched for are sold by someone else. For companies selling online, it is a signal that their product data may be indexed without explicit consent and used to redirect traffic to Amazon, which is a potential competitive risk to the visibility of their own online store.
Two audiences, two different impacts
What this means
For individuals
When shopping through AI assistants, product recommendations may increasingly rely on Amazon's index, even when the goods being searched for are sold by another store.
For a business
Companies selling online face the risk that their product data will be crawled and indexed by Amazon in order to redirect shopping traffic to its own pages, even for goods that Amazon itself does not offer.
Sales and marketingCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.