People talk about "AI search" as if it were a single thing you optimize for. It is not. On June 15, 2026 we asked the same 30 real questions a 6-to-8-figure ecommerce operator would type to three AI engines โ ChatGPT, Claude, and Perplexity โ in their live-web-search mode, and recorded which domains each one cited as sources. The engines disagreed almost completely. They pulled from different numbers of sources, favored different kinds of sites, and shared only a tiny sliver of common ground. This page lays out exactly how differently each engine cites โ every number comes from that run.
Methodology
- Run date
- June 15, 2026 (a single snapshot)
- Queries
- 30 real questions a 6-to-8-figure ecommerce operator would type
- AI engines
- ChatGPT (GPT-4o search), Claude, Perplexity โ each in live-web-search mode
- What we recorded
- Which domains each engine cited as sources across the 30 queries
- Counts
- Per-domain counts below are the number of the 30 queries that cited that domain
Honest limitations. Gemini was excluded from this comparison because its grounding citations did not parse reliably this run, so we leave it out rather than report numbers we cannot stand behind. And this is a single snapshot from one day, not a trend yet โ counts are out of 30 queries, and the right way to use these figures is as a baseline to re-measure against later.
"Optimize for AI Search" Is a Myth
Here is the headline finding, and it reframes the whole strategy: there is no single AI search. Across the same 30 queries, only 4 domains were cited by all three engines โ shopify.com, aiadvantageagency.com, searchengineland.com, and seo.com. That is the entire common ground between ChatGPT, Claude, and Perplexity. Everything else, each engine reached for on its own.
And the engines did not even pull from the same-sized pool. Counting the distinct domains each one cited across the identical 30 questions, Claude cited 213 distinct domains, Perplexity cited 166, and ChatGPT (GPT-4o search) cited 105. Claude pulled from roughly twice as many distinct domains as ChatGPT for the exact same set of questions. So "optimizing for AI search" is not one target โ it is at least three different targets, each with its own appetite, and they barely overlap.
There is no single AI search. Across the same 30 queries, only 4 domains were cited by all three engines, and the engines cited very different numbers of distinct domains โ Claude 213, Perplexity 166, ChatGPT 105.
Claude: SEO Agency and SaaS Blogs
Claude cited the widest field of any engine โ 213 distinct domains across the 30 queries โ and the sources at the top of its list lean toward SEO agency and SaaS blogs. shopify.com led at 7 of the 30 queries, followed by seo.ai at 5, then a band of agency and tool blogs. The observed pattern: Claude reaches broadly into specialist SEO and software content rather than concentrating on a few big names.
| Domain | Queries citing it |
|---|---|
| shopify.com | 7 |
| seo.ai | 5 |
| seoprofy.com | 3 |
| gofishdigital.com | 3 |
| outerboxdesign.com | 3 |
| nightwatch.io | 2 |
| contentpen.ai | 2 |
| ringly.io | 2 |
Perplexity: Heavy on YouTube and Social
Perplexity cited 166 distinct domains, and its top list looks nothing like Claude's. The standout: it cited youtube.com in 20 of the 30 queries โ by far the most concentrated single-source habit we observed in any engine. shopify.com followed at 9 queries, then a mix of established SEO tools and social platforms. The observed pattern: Perplexity leans heavily on video and social sources, pulling YouTube into two-thirds of the answers we measured.
| Domain | Queries citing it |
|---|---|
| youtube.com | 20 |
| shopify.com | 9 |
| ahrefs.com | 3 |
| semrush.com | 3 |
| google.com | 3 |
| facebook.com | 3 |
| eesel.ai | 3 |
| 1digitalagency.com | 3 |
ChatGPT: Tech and Mainstream Media
ChatGPT (GPT-4o search) cited the fewest distinct domains โ 105 โ and its top sources lean toward tech and mainstream media alongside SEO blogs. ecomseo.co and shopify.com tied at the top with 5 queries each, but the distinctive entries are TechRadar (4 queries) and The Atlantic (2) โ mainstream and tech press that the other engines did not surface at the top of their lists. The observed pattern: ChatGPT mixes specialist SEO content with established media brands, and concentrates on a narrower set of domains overall.
| Domain | Queries citing it |
|---|---|
| ecomseo.co | 5 |
| shopify.com | 5 |
| techradar.com | 4 |
| youtube.com | 4 |
| ahrefs.com | 2 |
| theatlantic.com | 2 |
| get-ryze.ai | 2 |
| searchflex.com | 2 |
Each engine has a different appetite: Claude reaches for SEO agency and SaaS blogs, Perplexity leans hard on YouTube and social (it cited YouTube in 20 of 30 queries), and ChatGPT mixes tech and mainstream media with SEO content.
What This Means for Your Store
Everything below is grounded only in the run above โ no assumptions, just what the numbers imply.
- You are not optimizing for "AI search" โ you are optimizing for several engines with different appetites. Only 4 domains were cited by all three. There is no single target to please; the same content lands differently in each engine.
- Diversify your source types to match where each engine reaches. Perplexity cited YouTube in 20 of 30 queries, so video gives you a path into Perplexity's answers. ChatGPT surfaced tech and mainstream media like TechRadar and The Atlantic, so earned media helps there. Claude leaned on SEO agency and SaaS blogs, so structured, specialist SEO content suits its appetite.
- Claude gives the widest field a shot; ChatGPT the narrowest. Claude cited 213 distinct domains versus ChatGPT's 105 โ so a broad-appetite engine offers more openings, while breaking into ChatGPT means clearing a higher bar against fewer slots.
- Re-measure before you conclude a trend. This is one snapshot from one day, with counts out of 30 queries. Treat these figures as a baseline and run the same queries again later to see what moved.
The deeper guides break down each engine's selection logic on its own: how ChatGPT Search picks sources, how Perplexity decides citations, and how Claude decides citations. For the full leaderboard across all engines, see the AI Citation Index.