Asking AI where to buy has become a habit. ChatGPT has more than 900 million weekly users and the Gemini app 950 million monthly. A growing share of product research and comparison now happens in these assistants instead of a search engine or a retailers’ sites.

For a retailer, that changes the goal. Being named in an answer is a first step, but it’s not enough. The shopper buys from the retailer whose products the assistant can best describe and match to the question: today's price, what is in stock, which model or size fits what the shopper asked for.

Shoppers who arrive from AI are worth more

Adobe, which tracks more than a trillion visits to US retail sites, found AI-referred traffic up 62% year on year in July, and these visitors convert 60% better than other traffic and bring 53% more revenue per visit. That makes sense: a shopper who has already researched and compared products in an assistant arrives knowing what they want, how the product will solve their issue, and is closer to buying than someone starting a search.

A mention is not a sale

Generative engine optimization (GEO) gets your brand cited in AI answers, and we think it is essential. But it only shows what an assistant could read about you on the web. It can get you named. It cannot tell a shopper today's price, whether the product is in stock or which promotion can apply. GEO and access to your live catalog work together: GEO gets you into the conversation, and the catalog gives the assistant something useful to say once you are there.

Even the web content an assistant can read is often incomplete. Adobe's AI Content Visibility Checker measures how much of a page's content an AI can read. Retail product pages scored 66% in Adobe's April analysis, against 75% for homepages, which means about a third of a product page's content cannot be read by an assistant. Many pages are also incomplete: information that is missing for human visitors is missing for an agent too, which then cannot tell whether a product fits the shopper's request. Product pages are where a shopper's question gets answered, so the lowest score sits on the page that matters most.

Give agents direct access to your catalog

A retailer has several ways to become agent ready, from cleaner product data to WebMCP. Another is a retail app built on the Model Context Protocol (MCP), the open standard that lets AI assistants connect to a company's systems. With it, ChatGPT or Claude can search your catalog with live prices and variants, check availability by store and warehouse, build a basket and track an order. Because MCP is open, the same app works in ChatGPT, Claude and other assistants that support it.

It also puts the brand in control of the conversation. Instead of an assistant paraphrasing the incomplete information it found about you, the shopper gets a shopping experience you designed, inside ChatGPT or Claude: your products, your prices, and the information you choose to expose.

Take a shopper asking for a smartphone around 450 euros. Without a catalog connection, the assistant answers from web pages that may be weeks old. With an MCP app, it lists the models you stock, with your product visuals and logo, at today's price, with availability near the shopper.

Retailers already in the conversation

At Alpic, we build the apps that let brands show up and sell inside ChatGPT, Claude and other assistants. Recommerce, Boulanger and Snipp are among the brands we work with, in retail and beyond. The same shift is reaching the large groups: we recently joined the Season 12 Pilot Program of La Maison des Startups LVMH, in the Omnichannel & Retail track, where the Group's Maisons pilot new solutions.

Test how ready you are

Before building anything, check what assistants can already say about your products. Ask ChatGPT, Gemini and Claude the questions your shoppers ask, and note where the answer is wrong, out of date or missing, or whether your brand appears at all. Those gaps show where to act. Improving the information on your own site and making sure agents can access it is work you can start today, in parallel with opening your catalog to assistants so they can read live prices, variants and availability. That second part is the one we help retailers with, putting you directly in the conversation.

Meet us at Tech for Retail

We will be at Tech for Retail at Paris Expo, Porte de Versailles, Hall 7.2, on November 30 and December 1, stand V77. If your test turned up gaps, contact us to set up time with our team at Tech for Retail.