Generative AI platforms fill 75% of their recommendation slots with the same handful of brands, and 78% of e-commerce brands never appear in AI recommendations in their own category at all. Those figures come from a new report by CommerceV3, an e-commerce platform that also sells AI visibility services. The company analyzed more than 1 million AI shopping prompts across ChatGPT, Gemini and Perplexity between March and July 2026 in the U.S. Each prompt ran five times per engine. A brand counted as visible only if it appeared in at least two of the five runs. In plain terms: for most brands, the AI answer box is a closed shop.

What did CommerceV3 measure?

The study covered seven e-commerce categories, about 200 subcategories and roughly 25,000 e-commerce brands, compiled from public marketplace listings, retail directories and category-level web commerce indexes. Example prompts included "best running shoes for flat feet" and "compare Brand A vs. Brand B."

The report finds that ChatGPT, Gemini and Perplexity frequently cite category leaders, marketplaces and editorially validated brands, while other brands go unmentioned. Each platform sources results differently, and the absence rates show it:

• Perplexity, which relies on live web citations: brands absent from 74% of category prompts
• ChatGPT, which leans on its training corpus and browsing: 76% brand absence rate
• Gemini, which draws on Google's index and Shopping Graph: 81% brand absence rate

Brands were most likely to be absent from problem-solving prompts (84%) and broad recommendation prompts (81%). Categories with strong editorial review ecosystems, such as consumer electronics and beauty, had a lower absence rate. Specialty and long-tail categories, such as hobby and grocery, saw much higher absence.

Why is AI absence worse than a poor Google ranking?

The report argues the stakes exceed a low search ranking. On Google, a brand ranking in the bottom half of page one, or on pages two or three, still picks up residual traffic. AI assistants offer no such floor. "An assistant recommends three to five options; every other brand receives nothing — no impression, no click, no signal it was even considered," the report states.

The consumer shift is measurable. In July 2026, visits to U.S. retail sites from AI sources rose 62% year over year, according to Adobe data cited in the report.

Where do the platforms get their answers?

The majority of AI recommendations point to third-party sources rather than brands themselves. The citation breakdown:

• Editorial and review sites — 34%
• Reddit and forums — 22%
• Marketplaces — 18%
• Brands' own sites — 14%
• Video content — 7%
• Other — 5%

Because generative AI pulls from brands' own websites only 14% of the time, brands need to make sure third-party sites are talking about them, the report concludes.

Can brands fix their visibility without ad spend?

The report says yes. Brands that AI cites share structural traits. Their content includes "extractable answer content" — concise, question-shaped pages that generative AI can lift answers from directly — plus schema markup that helps AI read the content. These brands also state clearly who they are and what they do, so AI does not rely on context clues, and they keep brand facts consistent across the internet. CommerceV3's takeaway: brands that invest time and talent can improve their AI visibility without a bigger ad budget. For the 78% currently invisible, that work is now the baseline, not the optional extra.