AI visibility for ecommerce: product cards in ChatGPT shopping

This guide defines how a store measures its product cards in AI answers. It reads the cards in the answers of our own project, and it lists what a seller needs and what a store tracks. Our project sells software, so its cards are few. They show what a card carries, not who wins a category. For the platforms beyond ChatGPT and whom each one serves, read our guide on which AI search engines matter. For the pages that get a brand named in the text of an answer, read how to get mentioned in ChatGPT answers.

What is AI visibility for ecommerce?

AI visibility for ecommerce is the share of AI shopping answers that show a product card of your brand. A shopping answer is an answer that shows at least one product card. A product card is a block inside the answer with a product's name, an image, a price and a link to a seller. Of the 7 platforms FullReach AI tracks, only ChatGPT reports its cards, so the measurement reads ChatGPT alone. The AI Shopping page shows the dashboard that holds it.

The cards are rare in a software category. In the 28 days to 30 October 2026, ChatGPT gave 1,307 answers to the 47 questions of our own project, in two markets. 9 of the 1,307 showed product cards, 0.7%. The 9 answers fell on days between 24 and 29 October 2026, and 8 of the 47 questions drew one. The question that drew most asked for the best tool to monitor Google AI Overviews, with 2 shopping answers of 28. Our reads cover this one software category only. A store first reads which brands the answers to its own product questions name.

What does a ChatGPT product card carry?

A card carries a product name, a price with its currency, the seller that offers it, and a link to that seller. In our 28 days, the 9 shopping answers showed 31 cards for 24 products. 20 of the 24 products carried a price: 16 in US dollars, 2 in Indian rupees, 1 in euros and 1 in pounds. 7 of the 24 carried a star rating. FullReach AI keeps a price in the currency the card shows, and never converts it.

The links show where the cards come from. We read the card links of one answer, from 29 October 2026. Three of its four card links carried the parameter utm_medium=feed, and its images carried Shopify product ids. These links suggest that a seller's product feed supplies a card, not the brand's own page. One answer is a sign, not a rule.

The cards do not always fit the question. 8 of the 24 products were physical goods with no link to software. They were two tape measures, a perfume, a sweatshirt, a women's suit, a back scrubber, a sports radar and a key finder. ChatGPT showed them beside software, for questions about software.

What does a seller need for its products to show?

Our reads support four items to different degrees, so each item names its evidence.

1. A product feed with a price and a currency for each market. Three of four card links in one answer came from a feed. Our cards showed prices in four currencies.
2. The brand name in every product title. A tool places a card by the spelling of its product name. The section on catalogs shows two cards that this spelling placed under the wrong brand.
3. Reviews on the product. 7 of the 24 products carried a star rating on their cards. Our reads do not show whether a rating moves a card up.
4. Your own store among the sellers. A card links the sellers that offer the product. FullReach AI counts the cards where your own store is one of them.

Our reads do not measure structured data on a product page, so this guide makes no claim about its effect on a card.

Which sellers do the cards send buyers to?

In our 28 days, the cards sent buyers to 33 sellers through 47 offers. In the same 28 days, one card for a monitored rival's software linked a software reseller's listing, not the rival's site. Another card sold a seat of an SEO suite through a marketing agency's store. The table lists the 12 sellers with two offers or more, 26 of the 47 offers. First offer counts the cards where that seller's offer comes first.

FullReach AI counts the offers per seller. It also counts your own store apart: the cards with an offer from your store, and those where that offer comes first. In our 28 days, both counts were 0, because no card showed our brand. For a store, the two counts say whether ChatGPT sends the buyer to you or to a reseller of your goods.

Why does a store need a catalog to count its cards?

A store needs a catalog because a product name alone can place a card under the wrong brand. Two cards in our 28 days show the trap. A card for a book carried a rival's name in its title, and the spelling rule placed it under that rival. A back scrubber with "FullReach" in its name went under Other brands. The AI Shopping page lists the catalog formats, and what the Products page shows once a catalog is in.

How do you measure which products and sellers win?

You measure the win with five counts over the shopping answers of one window. The table states each count and its value in our 28 days.

A card count and a mention count differ. In the answer of 29 October 2026, the text named four brands, and the cards showed five products from four sellers. The text mentions of the same answers follow the method in our guide on competitor presence in AI answers. The methodology states how FullReach AI collects the answers and reads the cards.

Which questions should a store track?

Track the questions a buyer asks before a purchase, one set per product kind, without your brand in them. A question names a product kind and a use, never a brand. Our project's 47 questions ask about software, and 8 of them drew a card in 28 days. A prompt is one question in one language, and each market counts once on the plan. Our guide on how many prompts to track gives the count. A one-person shop with no time for a daily read takes the monthly routine of our guide for freelancers and one-person firms.

A daily read of the cards runs on a plan. FullReach AI costs 65€, 165€ and 360€ a month, billed yearly, for 50, 150 and 350 prompts, with every platform included.

What to do next

1. Ask ChatGPT three product questions of your category in a private window. Note each card's product, price and seller.
2. Check your product feed against the four items of the seller list above.
3. Track the questions daily, and read first the shopping answers that show no card of yours.