Amazon has done this before. It built massive internal warehousing and logistics capability, then started selling those services to other businesses as Fulfillment by Amazon. It built its own cloud infrastructure to run Amazon.com, then turned it into AWS, which is now the world's largest cloud business. The pattern is consistent: build something that works at scale internally, then sell it to everyone else.
The latest version of that playbook is the Agentic Shopping Assistant on AWS. The product packages the technology behind Amazon's own AI shopping tool, called Alexa for Shopping, and makes it available to any retailer that wants to build a similar experience on their own website or app. Kate Spade, owned by the fashion group Tapestry, was the first to go live, launching what it calls an AI Gift Concierge in April 2026.
The Kate Spade assistant is a practical example of what this looks like in the real world. Instead of typing "handbags" into a search bar, a shopper can describe who they are buying for, the occasion, and their budget. The assistant then narrows down options through conversation. Tapestry says it built the assistant specifically around gift buying because that is where shoppers feel most lost. Amazon's own data shows 53% of shoppers report stress during gift purchases.
Amazon's pitch rests on two things: speed and proven results. Building a tool like this from scratch would normally take years. With the AWS package, which includes starter code, technical guidance, and access to Amazon's experts, retailers can be up and running in roughly 60 days. The underlying technology has been tested across billions of shopping interactions on Amazon.com, which Amazon describes as being "Customer Zero" for its own product.
The business case for conversational shopping tools is real, regardless of who builds them. Shoppers who engage with AI during a session convert at roughly four times the rate of those who just use search. Traffic to retail sites coming from AI sources grew nearly 4,700% year-over-year. The market for conversational commerce was worth around $26 billion in 2025 and is growing fast. Retailers that sit this out are not neutral, they are falling behind.
But there is a question that every non-Amazon retailer needs to answer honestly before signing up: do you trust Amazon with this layer of your business?
Amazon's retail arm competes directly with the brands that would use this tool. Tapestry's Kate Spade sells handbags. Amazon sells handbags. The same company now runs the AI engine deciding which Kate Spade products to recommend to shoppers on Kate Spade's own website. Amazon says retailers keep full control of their customer data and that the data does not flow to Amazon's retail side. AWS has operated this way for years, running infrastructure for Netflix, Walmart, and other Amazon competitors.
That separation has held up legally, but the discomfort among retailers is documented and longstanding. Microsoft has openly marketed itself to retailers as the cloud provider with no retail conflict of interest, and that pitch has worked. The question for retailers considering this tool is not whether Amazon is trustworthy in some abstract sense: it is whether the speed and capability of the AWS product outweighs the dependency it creates.
For retailers outside the fashion sector, the Kate Spade example still carries a lesson. Conversational shopping is most useful when the purchase decision is complicated: gifts, specialty products, things with many variations or options. A steel supplier selling industrial materials online would get less from this than a specialty retailer with a wide, nuanced catalogue. The use case fits best where shoppers already struggle to know exactly what to type into a search bar.
Google, Microsoft, OpenAI, and Walmart are all building competing versions of the same thing. This is not a uniquely Amazon product for long. The real decision is not whether to offer conversational shopping, but whose infrastructure you build it on.