From Chatbots to Agentic AI: A New Retail Shopping Layer
AI shopping assistants are conversational retail shopping agents that use generative AI and real-time catalog data to guide discovery, personalize product recommendations, and complete transactions for shoppers across digital and physical channels. Unlike static search bars or basic chatbots, these agentic AI systems can interpret intent, ask clarifying questions, and coordinate multiple actions such as search, comparison, and checkout in a single dialogue. As shoppers grow used to AI-guided journeys, retailers are racing to build their own assistants that reflect their brand voice and product expertise. This shift is not only about answering questions faster; it is about building continuous, personalized journeys that can stretch from first inspiration through to post-purchase support, while feeding back data that improves merchandising, marketing, and overall AI customer experience over time.
Amazon’s Agentic Shopping Assistant Becomes a Retail Product
Amazon is turning its internal AI shopping know-how into a commercial service with the Agentic Shopping Assistant on AWS, giving retailers a shortcut to their own branded agents. Built on Amazon Bedrock, AgentCore, and OpenSearch, the solution is designed so merchants can deploy conversational agents in weeks instead of years, then adapt them to their catalog, customer base, and tone of voice. Kate Spade New York is an early adopter, introducing an AI Gift Concierge that recommends items by occasion and style through natural-language conversations, based on real customer feedback about the stress of gift shopping. Amazon says its AI shopping assistant on Amazon.com served more than 300 million customers last year and generated nearly USD 12 billion (approx. RM55.2 billion) in incremental sales, which now forms the data-tested foundation for third-party retailers that adopt the new agentic AI blueprint.
Kmart’s Joy Shows What Immersive AI Customer Experience Looks Like
Kmart’s Joy illustrates how AI shopping assistants are moving beyond text chat into interactive, visual experiences. Built with Google Cloud’s Gemini Enterprise for Customer Experience, Joy sits inside the retailer’s website and app as a conversational guide across its full marketplace range. Shoppers can describe what they need in natural language—size, style, color, and budget—and receive targeted, personalized product recommendations in seconds. Visual tools sit alongside this: virtual try-on helps people see how selected products might look on them, while the “See It in My Space” feature overlays furniture or décor into their homes before purchase. Customers can also upload photos to get tailored suggestions and side‑by‑side comparisons. Because Joy can span the entire customer lifecycle, from discovery to post‑purchase support, it points toward a future where retail shopping agents provide an always‑on, context‑aware AI customer experience rather than isolated one‑off interactions.
Agentic AI Inside Shopify: Nosto and Reddit Close the Loop
While front‑end assistants guide shoppers, back‑end retail shopping agents are reshaping how e‑commerce teams manage discovery and marketing. Nosto’s integration with Shopify Sidekick brings agentic AI workflows directly into the Shopify admin, so merchandisers can describe changes they want in plain language and see Nosto-powered product recommendations update on their storefront without switching tools or relying on developers. According to Nosto’s CEO Jim Lofgren, “the bottleneck for most commerce teams is operational bandwidth and access to developer resources,” and agentic AI inside Sidekick aims to remove that. On the acquisition side, Reddit’s expanded Shopify integration automates product catalog syncing and conversion tracking for Dynamic Product Ads, connecting product discussions to live inventory. Reddit cites TransUnion research showing its campaigns in North America delivered more than double the incremental ROAS of the average media plan, with an average seven-times ROAS for retail advertisers in EMEA.
Toward Autonomous Retail Shopping Agents
Taken together, these launches signal a shift toward autonomous retail shopping agents that can manage entire journeys: discovering products, delivering personalized product recommendations, and driving e‑commerce automation right through to checkout. Amazon’s AWS offering provides a battle-tested technical foundation, Kmart’s Joy shows how immersive multimodal experiences can feel, and Nosto plus Reddit highlight how agentic AI will also handle merchandising optimization and ad performance in the background. Over time, these systems are likely to coordinate across channels, remembering preferences from an AI shopping assistant on a site, then retrieving them when a shopper asks a voice agent in-store. For retailers, the challenge is no longer whether to adopt AI shopping assistants, but how to design agents that respect brand promises, protect data, and balance automation with human oversight while still delivering a seamless, high‑performing AI customer experience.
