TL;DR

• ChatGPT's product carousels don't run on some mysterious AI index.
• They pull directly from Google Shopping's top 40 organic results and apply their own re-ranking on top.
• If your products aren't there, they don't exist in ChatGPT's recommendations.
• The good news: if you already have a Google Merchant Center feed, you're 80% of the way there.
• Here's everything we covered in our second Precis Masterclass, with Peec AI's Malte Landwehr and Precis SEO & AEO Director Kristina Bergwall.

Why we dedicated a full masterclass to ChatGPT Shopping

It's a fair question. But the numbers make a strong case.

OpenAI officially reports 159 million monthly active search users in the EU alone — more than the combined populations of Sweden, Germany and Italy. And ChatGPT isn't just one option in a crowded market: it holds 76% of the EU AI chatbot market share (Statcounter, May 2026), with Google Gemini at 8.7% and Microsoft Copilot at 5.7%. For now, optimising for ChatGPT is effectively optimising for AI search in Europe.

What those users are doing inside ChatGPT matters just as much. According to McKinsey's Consumer AI Discovery Survey (December, 2025), 63% of European consumers already use AI to compare brands, models, prices and reviews. 46% use it to learn about a category or product. Another 46% use it to discover or get inspiration for purchases. These aren't early adopters experimenting with a new tool — this is mainstream consumer behaviour, happening now.

And product carousels show up more often than you might expect. For apparel and fashion, they're triggered 62% of the time. For other physical products, 56%. Even for groceries and consumer goods, it's over 30%. When you run the numbers (159 million users, 46% using AI to research products, a conservative estimate of 3 product-related searches per month, and a 62% carousel trigger rate for fashion), you get approximately 136 million potential product carousels shown to European consumers every month.

That's not a niche feature. That's a whole channel.

The finding: ChatGPT Shopping is scraped Google Shopping

Here's what Peec AI's research uncovered, and it's more straightforward than almost anyone in the industry expected.

When a user asks ChatGPT something that triggers a product carousel (ie: "best running shoes under €150," "what espresso machine should I buy"), ChatGPT doesn't search its own product database. It fires off what Peec AI calls shopping fanout queries: short, product-focused search queries sent directly to Google Shopping's organic results. The products that come back from those queries are what populate the carousel.

The proof is in ChatGPT's own source code. Peec AI researchers found a hidden encoded field in the page, and when decoded, it revealed unmistakable Google Shopping parameters: product IDs, catalogue IDs, offer document IDs, locale parameters, and the exact query used to retrieve the product. They were then able to reconstruct a working Google Shopping URL from those parameters — and it matched the product in the carousel exactly.

Malte Landwehr tested this personally. His wife launched a food brand selling sugar-free cookies. One day after connecting the product to Google Merchant Center, it appeared in ChatGPT Shopping with the same prices, availability data and merchant links pulled straight from the feed. Without a partnership with OpenAI. Without separate integration. All based on a Google Merchant Center feed.

The scale of the research confirms what the anecdote suggests. Peec AI analysed over one million shopping fanout queries, compared against Google Shopping results across many industries and markets. The conclusion: 100% of products seen in ChatGPT Shopping can be explained by the top 40 organic products in Google Shopping for the corresponding query. Bing Shopping, by contrast, explains only 11%... and almost none of those exclusively. If it's in ChatGPT's carousel, it came from Google.

What "organic only" actually means here

This is worth pausing on, because it changes the commercial logic significantly.

ChatGPT looks at organic Google Shopping results only. Paid ads are completely ignored. That means appearing in ChatGPT's product carousels costs nothing beyond the work of having a well-optimised Google Merchant Center feed. No platform fees, no sponsored placements, no new budget line. It's a free, high-converting channel that most e-commerce brands are already most of the way equipped to access.

The conversion rate data reinforces why this matters. According to Adobe for Business (March, 2026), LLM traffic converts 42% higher than non-AI traffic. The reason is intuitive: someone asking ChatGPT for a specific product recommendation isn't browsing — they're in a high-intent purchase mindset, looking for a definitive answer. When your product shows up in that moment, the path to conversion is short.

How ChatGPT decides which products to show

It's not a straight copy of Google Shopping rankings. ChatGPT applies its own re-ranking on top of what Google returns, which is where the real opportunity (and nuance) sits. A few things Peec AI's research reveals about how that re-ranking works:

Position matters, but it's not everything

The average difference between a product's Google Shopping position and its ChatGPT carousel position is about five places. Products get reshuffled. That means ranking first in Google Shopping doesn't guarantee appearing first in ChatGPT, and ranking fifth doesn't rule you out of the top carousel slot.

Product attributes signal relevance

When ChatGPT generates a comparison table, it rates products on attributes like comfort, durability, versatility, or price range. These attributes aren't random — they reflect what ChatGPT considers the most important dimensions of that product category. If your product description doesn't use those words in a positive, evidence-backed way, you're at a disadvantage in the re-ranking process.

Star ratings act as a filter

‍Products with fewer than three stars on the attributes ChatGPT evaluates are effectively invisible in the carousel. Proving product quality in your feed and on your product pages isn't just good practice — it's a threshold requirement.

Shopping fanout queries introduce new terms

‍This is one of the most actionable findings in the research. ChatGPT's shopping fanout queries often include words that weren't in the user's original prompt. In Malte's example, a user asked for a "cozy, heavyweight, oversized hoodie" — and the fanout query added "premium cotton," "structured," and "men." If those terms don't appear on your product detail page or in your feed, your product is less likely to be retrieved in the first place.

The top terms ChatGPT adds to fanout queries across e-commerce include: "best", "review(s)", "2026", "top", "comparison", and "vs". These aren't just keywords — they're a direct signal of the content types that should exist around your products.

Global versus local: it depends on the category

One finding that senior marketers operating across multiple European markets should note: ChatGPT's product recommendations are not uniformly global or local — they vary significantly by category.

For skincare, the same three brands appear consistently across the US, Mexico, Germany, Austria, Switzerland, Spain, Sweden and the UK. A globally strong brand presence in a category with established consensus translates directly into consistent AI visibility across markets.

For wedding outfits, the picture is almost entirely different by country. ASOS dominates across several European markets. Bubbleroom and NA-KD appear in Sweden. Lulus and Nordstrom appear in the US. The brands ChatGPT recommends reflect the local Google Shopping results for each market.

The implication for marketers: in commoditised or consensus-driven categories, global brand authority matters more. In fashion, occasion wear, and other culturally specific categories, local market presence and local feed optimisation are what determine visibility.

The bottom line

ChatGPT Shopping is not a new, separate channel that requires a new, separate strategy. It's Google Shopping's organic results, re-ranked by an AI layer. The brands already doing the fundamentals well — clean feeds, strong product data, good ratings, rich descriptions — are closer to AI visibility than they think.

Upgrading your organic feed for AI visibility is essentially a free, high-converting revenue channel that you are already 80% equipped to handle. Kristina Bergwall's three-level framework (feed fundamentals, feed enrichment, and conversational domination) covers the remaining 20%, starting from wherever you are right now.

👉 Read the framework post: How to show up in ChatGPT Shopping: a three-level playbook for e-commerce brands

👉 Want to understand your current AI visibility and where to start? Get in touch with the Precis team.