By Carson Cobb. Data collected November 4 to 17, 2026.
I asked ChatGPT, Gemini and Google AI Mode 48 shopping questions, eight in each of six product areas, and recorded every brand they named. Then I scanned 617 established Shopify brand stores in the same six areas and checked which of them came up, and what the named ones do differently.
The short version
- 65 of 617 brands were named at least once. 11%, across 144 answers.
- The answers named 391 different brands in total. Most aren’t in my dataset, which only includes brands with their own Shopify store, so many are larger brands sold through retailers or other platforms.
- Named brands publish much more. A median of 142 blog articles versus 66.
- And more recently. 65% posted in the last 90 days, versus 50%.
- Their product pages say more about proof. 40% mention testing or certification versus 29%, and 34% compare options versus 24%.
- Their structured data isn’t better. Product structured data was on 87% versus 86%.
Named versus not named
| Measure | Named by AI (65) | Not named (552) |
|---|---|---|
| Median blog articles | 142 | 66 |
| Blog with 20 or more articles | 74% | 67% |
| Posted in the last 90 days | 65% | 50% |
| Product page mentions testing or certification | 40% | 29% |
| Product page compares options | 34% | 24% |
| Median words on the product page | 1,375 | 1,131 |
| Product structured data | 87% | 86% |
| GTIN in structured data | 30% | 22% |
| Organization data with sameAs links | 75% | 79% |
| Two or more Product blocks | 19% | 21% |
| Median product description words | 56 | 62 |
How to read this
These brands didn’t get named because of their blog. AI mostly names brands that independent sources already write about. In my earlier research, ChatGPT showed sources in 45 of 48 answers, led by review and testing sites and magazines (sources ChatGPT cites).
What the pattern shows is the kind of brand those sources write about: one that publishes a lot, keeps publishing, and backs claims with testing and comparisons. That’s material a reviewer can use. The technical layer, schema and identifiers, was about even. It matters for getting facts right once AI is considering you, as I found in what AI gets wrong about products, but it didn’t separate who got named.
What to do with this
- Give reviewers something to cite. Test results, certifications, comparisons and clear specs on the product page.
- Publish steadily, on buyer questions. “Which one is right for sensitive skin” beats brand news.
- Earn third party coverage. Review sites, testing publications and gift guides are what AI reads most.
- Then fix the data. Structured data, GTINs and policies make sure AI repeats the right facts.
Finding out where AI names you and why is part of my AI Search and Conversion Audit. The AI Visibility Check is a quick first look.
How this was done
- AI answers: 48 shopping questions, eight each in supplements, beauty, food and drink, pet, home and apparel, asked in ChatGPT, Gemini and Google AI Mode between November 4 and 11, 2026. Every brand named in each answer was recorded.
- Stores: 617 Shopify brand stores in the same six product areas, my full dataset of 620 minus client stores and one store that blocked automated reads, scanned November 16 and 17, 2026.
- Matching: brand names matched to store domains, then checked by hand. Brands with a common word as a name were checked against the question they appeared in.
- Limits: eight questions per area can’t cover every product a store sells, so an unnamed brand may simply not fit the questions asked. This shows correlation, not cause.
No store or brand is named in this report.