Method
How a number gets printed
Every number on this site carries an engine, a route, a repeat count and a date. This page is where those words are defined, so a client or a reader can check our work rather than take it on faith.
Engines and routes
Scroll for more columns →
| Engine | Production route | This census used |
|---|---|---|
| Claude | Messages API (next census) | Claude Code CLI with web search, Sonnet, 3 repeats per prompt (this research pass) |
| Gemini | Gemini API | Not yet in the public census |
| Google AI Mode | Disclosed data vendor | Not yet in the public census |
| ChatGPT | Responses API, web search tool | Not yet in the public census |
| Perplexity | Hand checks today, API route planned | Not yet in the public census |
The one route published in this first census, the Claude Code CLI with web search, is Anthropic's own tool, not a scraper. Gemini and Google AI Mode answers exist from the same research pass, but they came from scripted browser sessions in signed-in accounts, which Google's own systems rate-limited as unusual traffic after about 76 queries; that is not a route we publish numbers from. The next census moves Claude to the Messages API and adds Google through a disclosed data vendor.
Repeats and intervals
One run is one draw from how a model answers that day, not a fact about the brand. A brand at a 50% mention rate still carries about ±20 points of uncertainty after 20 repeats. The square strip below is a worked, illustrative example of what that looks like at the repeat level, not a real brand.
Census tables print a cluster-bootstrap range as the headline figure, since it accounts for prompts repeating, not just answers, which widens it. A narrower Wilson range sits beside it, labelled: Wilson treats every answer as an independent draw, which is the more optimistic of the two, so we do not lead with it.
Illustrative: 10 of 20 repeats, not a real brand
Two 2026 papers argue for bootstrap intervals and adaptive sampling on exactly this problem. No small-business-priced tool prints the interval at all; we print it on every table on this site.
The source-integrity check
Before a domain an engine cites counts toward anything on this site, it gets three checks: its Tranco popularity rank, whether its homepage actually matches the category it is being cited for, and whether it shows signs of a recent ownership change (a domain that changed hands and now hosts something unrelated to its history). A domain that fails is marked FLAGGED in text, never by colour. A domain that passes is marked VERIFIED. A domain we have not yet checked is marked UNRANKED.
Scroll for more columns →
| Domain | Cites | By route | Created | Tranco rank | Status |
|---|---|---|---|---|---|
| Wedding-venue domain, name withheld | 47 | Claude CLI 46 (Sonnet 39, Opus 4, Haiku 3) · claude.ai app 1 · Google AI Mode 0 | 1997 | Not ranked | FLAGGED |
Roughly one in seven of the Claude Code CLI's citation instances in our research got flagged for a human look this way; on a human read, about a third of flags turned out to be on-topic after all. That precision figure stays internal and out of client-facing numbers until the flags are hand-labelled at scale, which is also why this page, not a client report, is where the caveat lives.
What we will not promise
Every mention rate on this site carries a route, a repeat count and a date, which is the discipline this page exists to enforce. That discipline rules out the shortcuts that make a number easy to sell before it is earned: a ranking promised in advance, a point score with no repeat count behind it, or a consensus that was manufactured (bought reviews, hidden crawler text, incentivised posts) rather than measured. A number that cannot carry its route and its count does not get printed, on this page or anywhere else on the site.
Which bots can read this site
robots.txt here allows the major search and AI crawlers: Googlebot, Bingbot, OAI-SearchBot, GPTBot, ClaudeBot, Claude-SearchBot, PerplexityBot and Google-Extended. Google-Extended specifically controls whether Google's Gemini app can train on and ground its answers in this content. It does not control whether this site shows up in Google Search or in AI Overviews, which Googlebot governs on its own. We leave it allowed anyway: a census nobody can read is not a census, and the trade-off is that anything crawlable here can also end up reused by a model we have no contract with.