Retailers have found a source of shoppers: the chatbot window.
On August 7, 2026, Reuters reported that Walmart, Ulta Beauty, Wayfair, Etsy, Kate Spade and The Knot are reworking digital storefronts so products appear more often in ChatGPT and Google Gemini recommendations. Brands want AI to discover products, but many still want shoppers to finish purchases on the retailer’s own site.
That split is becoming the central economic question in AI commerce. Chatbots can deliver qualified buyers, yet checkout controls the customer relationship. Checkout reveals basket size, purchase frequency, loyalty activity and product preferences.
The shift echoes AbacusNews.com’s coverage of the AI shopping shift: discovery is moving into conversation, but transaction data remains the prize.
AI Referrals Are Becoming High-Value Traffic
Adobe Analytics said 41% of U.S. consumers used generative AI for online shopping in June 2026. Visitors referred by AI services produced 41% higher revenue per visit than shoppers coming through traditional channels. Juniper Research expects shoppers to spend $8 billion this year after AI agents direct them to retail sites.
Shopify sees the same pattern. Its Q2 2026 results showed AI-driven traffic and orders to merchant stores nearly tripled from a year earlier. President Harley Finkelstein said AI complements search by matching buyer intent against structured product data.
Shopify’s AI commerce data found AI-referred shoppers convert at nearly 50% higher rates than organic-search visitors on product pages and carry 14% higher average order values.
Ulta Wants Discovery Without Surrendering Checkout
Digital and e-commerce chief Josh Friedman said Ulta is seeing roughly double the conversion and intent from shoppers finding products through Gemini and ChatGPT. The retailer is working with Google to bring carts and Ulta Beauty Rewards into AI-supported shopping.
Ulta still prefers customers to complete transactions on its website.
That preference protects more than payment. Loyalty data lets Ulta connect purchases across time, identify preferred brands and personalize offers. An AI platform may know what a shopper asked today. Ulta can know what that shopper bought for years.
Amazon Web Services digital commerce chief Vince Koh told Reuters that retailers retain an advantage through direct customer relationships. AWS encourages clients, including Kate Spade, to use AI platforms for discovery yet keep their websites as the best place to buy.
Structured Product Data Becomes The New Shelf Space
Chatbot visibility requires a different form of merchandising. AI shopping agents can evaluate dimensions, materials, inventory, price, compatibility, ratings and use cases across many product records at once.
Shopify says half of AI-referred sessions now land directly on product-description pages, 2.5 times the rate seen in traditional search. Retailers are cleaning catalogs, adding structured attributes and making inventory easier for machines to interpret.
Google’s retail AI agents push this model further, giving brands tools to build their own assistants instead of handing the customer experience to third-party chatbots.
First-Party Data Is The Defensive Wall
The fight is no longer about whether retailers participate in AI commerce. It is about where the relationship becomes owned.
A chatbot can introduce the shopper. The retailer wants account creation, loyalty enrollment, checkout, returns, support history and repeat purchase. Those events create first-party data that improves recommendations and raises lifetime value.
Etsy sees that pattern already. Chief Product and Technology Officer Rafe Colburn said shoppers who discover products through ChatGPT usually return to Etsy to complete the transaction. The referral starts the relationship; Etsy wants the next purchase to begin inside Etsy.
The Knot follows a similar playbook, optimizing vendors for ChatGPT discovery but steering couples back to its site for venue and invitation bookings.
OpenAI ended Instant Checkout in March 2026 and shifted toward product discovery with merchant-owned checkout, reducing fear that chatbot platforms would absorb the transaction layer.
Retailers Are Building A Two-Layer Commerce Model
AI commerce is splitting retail into two layers: discovery and ownership. ChatGPT, Gemini, Claude and other agents interpret intent and compare products. Retailers control checkout, loyalty, service and purchase history.
Brands that win both layers can gain high-intent referral traffic without handing valuable customer records to an intermediary. Brands that lose ownership risk becoming fulfillment engines behind an AI interface.
The referral boom is creating a new bargain: let chatbots send the shopper, but keep the customer. The sale matters. The data attached to the sale may matter more.
