AI search optimization for ecommerce stores.

Shoppers now ask AI assistants which product to buy, and the answer names a few products and the stores that sell them. Those visits are growing fast and convert well. Here is where the answers get their products, and what a store can fix.

Why it matters now

The visits are small, and growing fast.

Adobe measured more than 1 trillion visits to US retail sites. Visits from AI assistants are still a small share, but they grew fastest and bought more.

393%more visits from AI assistants to US retail sites in the first quarter of 2026 than a year beforeAdobe data, reported by CMOtech, April 2026
42%better conversion for visits from AI assistants than for other visits, in March 2026Adobe data, reported by CMOtech, April 2026
66%of a retail product page's content can be read by AI on average, the lowest of any page type Adobe measuredAdobe data, reported by CMOtech, April 2026
Where the products come from

3 places a shopping answer reads.

Product feeds

ChatGPT takes product data, such as price and availability, from feeds that merchants share with OpenAI (OpenAI's commerce docs). On Google, the product data you keep in Merchant Center matters too.

Your product pages

Price, sizes, stock, materials, delivery and returns, written in the page's text. Details that appear only in images or load after the page may never reach the assistant.

Reviews and buying guides

Review sites, forums and buying guides that name your products. An assistant leans on them to explain why one product suits a shopper better than another.

What to fix first

4 changes most stores need.

01

Share your product data

Send a product feed where the assistants take one, and keep prices and stock current in it.

02

Write the facts in the page

Put what shoppers compare, such as price, size, materials, delivery and returns, in plain text on each product page.

03

Answer the buying questions

A short guide for each question shoppers ask, such as which running shoe suits flat feet, with your products in it and the reasons why.

04

Correct old prices at the source

An old price on a comparison site can end up in an answer. Ask the site to fix it, and fix your own old pages.

How Visibian helps a store

See which products the answers name.

Add the questions your shoppers ask, such as the best product for a need or a budget. Visibian asks all 7 assistants and shows which products and stores each answer names, and the pages it cites.

Hallucination Watch flags a wrong price, size or return policy. On Business and agency plans, Ground Truth shows the real visits AI answers sent you, from your Google Analytics or your site's edge.

Questions

Asked before you ask.

How do I get my products recommended by ChatGPT?

Share a product feed with OpenAI, put the facts shoppers compare in your product pages' text, and earn mentions on the review sites ChatGPT reads. Then check which questions name your products.

Does generative engine optimization work for Shopify stores?

Yes. The steps are the same on any platform: product data the assistants can take, product pages they can read, and mentions on sites they trust. Feed apps exist for most platforms, Shopify included.

Are visits from AI assistants worth it for a store?

Adobe's data says they convert well: in March 2026, visits from AI assistants to US retail sites converted 42% better than other visits. Your own numbers can differ, and Ground Truth shows them.

Does Visibian track Google AI Overviews?

No. It tracks Google AI Mode, alongside ChatGPT, Google Gemini, Perplexity, Microsoft Copilot, Grok and Claude.

Get started

See what AI tells your buyers about you.

Plans from €79 a month before VAT, with all 7 assistants on every plan. Or start with the free AI Readiness Check.

Subscribe

For online stores: why an assistant names other products and leaves you out.