Amazon is offering other retail companies the use of its AI shopping assistant on their websites. It will be interesting to see who actually has enough confidence to run their most sensitive data on the technology of their biggest competitor.
AWS offers the Agentic Shopping Assistant as a solution enabling retail companies to build their own dialogue-based shopping assistants for their shops. The technology draws on experience from Alexa for Shopping, the successor to Rufus.
The new service is designed to provide retail companies with architecture, starter code and support from AWS experts. Amazon promises that companies will be able to launch their own shopping assistants in weeks rather than years. The solution is already listed on the AWS Marketplace, but AWS only provides prices there on request.
US luxury retailer Tapestry uses Amazon technology
Kate Spade, the luxury handbag brand owned by US luxury fashion group Tapestry, is the first publicly named retailer to have launched such an application. The parent company Tapestry introduced the Kate Spade AI Gift Concierge back in mid-April. The assistant asks for the occasion, the recipient and the style, and suggests suitable products based on this.
Amazon markets the service as a way to keep the customer relationship within the retailer’s own shop. AWS states that retailers can integrate the solution with their own data, business rules and brand voice. Each implementation is designed to be tailored to the product range, target audience and shopping environment of the respective company.
Highly sensitive data
Yet this is precisely where the offer’s sticking point lies. Other retailers are likely to have little interest in seeing sensitive product range, customer or demand information end up with such a dominant competitor. Whilst AWS states that companies retain control over customer data, catalogues and business rules, no publicly verifiable details have been provided on how Amazon Retail and AWS organisationally and technically separate this data within this specific product.
The concern is not abstract. TechCrunch had already reported in March that Amazon was integrating product feeds from other retailers into its search and into its then-assistant Rufus via Shop Direct. The article pointed out that such data could give Amazon insights into popular brands, products and price points.
Amazon argues on the basis of scale and speed
Amazon cites its own scale as a selling point. According to the company, more than 300 million customers used the Amazon Shopping Assistant last year. Furthermore, the tool generated nearly 20 billion US-Dollars in additional revenue the previous year.
This presents a dilemma for retailers. The AWS service can accelerate the development of a proprietary AI assistant and is based on shopping data that Amazon has collected on a massive scale. At the same time, companies must decide whether they are willing to accept Amazon, of all companies, as a technical partner for a customer-facing AI layer.
Initial results are still pending
It is not yet known which other companies, apart from Tapestry’s Kate Spade, are testing the product. Amazon only refers to additional retailers in the test phase. Furthermore, Kate Spade’s initial KPIs have not yet been published.
