This page exists so you can check our work. It states what we actually ask, how many times we ask it, how wide the uncertainty is, and — the part most tools leave out — what we cannot see at all.
If a figure isn't on this page, we don't publish it.
Every visibility figure we show comes from an answer a model actually gave to a question we actually asked, stored with its full text so it can be re-read later. There is no extrapolation from clickstream data, no modelled search volume, no inferred "AI rank".
The cost of that rule is that our numbers are sometimes smaller, and sometimes absent. A brand with too few answers to be sure about shows not enough data rather than a confident-looking percentage. We think a visible gap is more useful than a number you can't lean on.
A probe is one prompt, sent to one model, once. The answer is scored for three things: whether the brand was named, how favourably it was described, and which sources the model cited. The full response text is stored alongside the score.
The headline figures pool engines. The Mention Rate, the sentiment figure and share of voice each count the answers of every engine in scope as one number, and the overview's combined score is built from them. Two engines agree on which brands to name far less often than people assume, so each engine also has its own rate on the per-engine panels — look there before acting on a single number.
| Provider | Model probed | We call it | Answers collected | Most recent |
|---|---|---|---|---|
| OpenAI | gpt-4o-mini | ChatGPT | 7,962 | 29 Sep 2026 |
| Anthropic | claude-haiku-4-5-20251001 | Claude | 6,072 | 29 Sep 2026 |
| gemini-2.5-flash-lite | Gemini | 4,777 | 29 Sep 2026 |
These are the exact model identifiers, not marketing names. "Claude" here means Claude Haiku 4.5 through the API — not the Claude consumer app, which has its own system prompt and retrieval behaviour and can answer differently. Web search is off by default (it is opt-in per run) and is recorded on every answer, because it materially changes what a model cites.
| What | Count | Read it as |
|---|---|---|
| Answers collected | 22,454 | Every stored answer, all engines, since we started: 18,811 from the three engines we track and 3,643 from other surfaces: Google AI Overviews and AI Mode, read from Google's results page, and the engines run only once (section 06). Nothing is estimated into this number. |
| Period | 27 May – 30 Sep 2026 | A 127-day span, but answers were collected on 41 separate days inside it. We probe in rounds, not continuously, and we do not call that "daily". |
| Engines | 3 | ChatGPT, Claude and Gemini. What we do not read is section 06. |
| Brands | 23 | Each with its own prompt set and competitor list. Some of them are industry-study rosters rather than client brands. |
| Prompts probed | 1,496 | Distinct questions asked at least once. 2,771 are configured; we count the ones that have actually produced an answer. |
| Markets | 9 | Country-tagged prompt sets. What "measured in a market" is worth depends on the language the answers came back in — section 05. |
| Citations recorded | 53,891 | Source URLs the engines themselves returned. |
| Answers with web search on | 7,214 | 32% of the corpus. Search is opt-in per run and costs roughly 17× an ungrounded answer, so most rounds are ungrounded. Grounded and ungrounded answers are not kept apart yet: a mention rate on the dashboard counts both together, so one engine's rate can blend answers that searched the web with answers from the model's memory. |
This is a small corpus next to the industry studies we cite elsewhere, and we would rather say so than pad it. Every one of these answers exists as stored text.
Claude answers stopped on 19 August 2026 and resumed on 10 September 2026. Our own Anthropic account had reached the spend limit set on it. Rounds run inside that gap were two-engine rounds, so a Claude figure covering those weeks rests on fewer rounds than the ChatGPT and Gemini figures beside it. This was our billing rather than a fault in the product. The latest Claude answer is from 29 September 2026.
Language models are not deterministic. Ask the same question twice and you can get two different brand lists. That means a mention rate is a sample statistic, and a sample statistic without an interval is a guess wearing a suit.
The headline Mention Rate and the buying-stage matrix show a 95% range grouped by question. The answers to one question move together, so asking it again is not new evidence and cannot narrow the range. The range is never narrower than the Wilson score interval over the individual answers (z = 1.96), which stays sensible at the edges — at 0 hits out of 12, or at small n.
The other rates on the dashboard — per engine, per competitor, per topic, per market — do not show a range yet.
| Rule | Value | Why |
|---|---|---|
| Minimum sample | n ≥ 7 | Below 7 answers the headline Mention Rate and the buying-stage matrix show no rate — they read "n/a" or "not enough data". A small n is not a zero; it means we haven't looked hard enough yet. Other panels do not apply this floor yet. |
| Interval | 95%, grouped by question | Shown beside the headline Mention Rate and in the buying-stage matrix, never narrower than the Wilson interval over answers. Two rates whose ranges overlap are not a ranking; some panels still order rates without that check, for example the strongest and weakest engine. |
| Precision | ±10pp ≈ 100 runs | Roughly what it takes to pin a mid-range rate to ten points. Tighter answers need more runs, and more runs cost money — we'd rather tell you the width than hide it. |
Measured counted directly from stored answers · Derived arithmetic on measured values, no new assumptions · Not offered we don't publish it
| Metric | Basis | How it's computed |
|---|---|---|
| Mention rate | Measured | Answers naming the brand ÷ answers collected, per engine. The headline rate and the buying-stage matrix carry a 95% range grouped by question; the per-engine figures do not show one yet. Branded prompts (where the brand is named in the question) are reported separately — averaging them in is how a tool tells you 62% when the honest earned figure is 0%. |
| Sentiment | Measured | Scored from the answer text where the brand appears, and only on answers that actually mention the brand. The scorer reads English; on an answer it cannot read it abstains rather than returning neutral, because a 0.0 nobody wrote is a fabricated data point. That currently silences 3,284 of 9,702 brand-mentioning answers (34%): 34 of 6,426 English ones — but every one of the 2,062 French, 313 German, 277 Portuguese, 246 Spanish, 204 Italian, 148 Polish ones. Today that means we report sentiment only for English answers; for every market answered in its own language there is none. |
| Citations | Measured | URLs the model itself returned. We never infer a citation from the fact that a page ranks in Google. |
| Share of voice | Derived | Your mentions ÷ all tracked-brand mentions on the same prompt set. Only comparable within one prompt set and one engine — it is not a market share. |
| Website readiness | Derived | Rule checks against pages we could actually fetch. A site we could not read returns no score at all — the field is left empty and labelled unread, never given a low grade. A bot wall tells you nothing about the content behind it, and grading it anyway would invent a finding. |
| Prompt volume | Not offered | Nobody can currently observe how often real people ask an assistant a given question. Tools that publish this figure derive it from clickstream panels and say so in their own documentation. We'd rather leave the field blank. |
| "AI rank" | Not offered | There is no ranked results page to hold a position in. We report how often you're named, not what position you hold, because the second thing doesn't exist. |
A market is a country. A country is not a language, and an assistant asked a localized question will often answer in English anyway. If nobody checks, an English answer counted into a Polish market's number turns into "your Polish visibility is 38%" — a sentence about Poland built from answers that were never Polish.
So every stored answer now carries the language it was actually written in, detected from the answer text rather than assumed from the question. Where a market's answers came back in another language, the number on screen says so, next to itself, with the share. It is a statement of what was measured, not a warning: these are real answers that really mention the brand — the only thing that was ever wrong is what we called them.
Once the same question has been re-asked on the same engine and answered in the market's own language, the earlier off-language answer is excluded from that market's figure and stays in the database. Same question and same engine, deliberately: dropping every off-language answer the moment any local one exists would shrink some markets to one or two answers, and a "market figure" computed from one answer hides that we measured at all.
| Market | Answers | In a local language | What that means |
|---|---|---|---|
| Spain | 493 | 100% | Fully measured in Spanish. |
| United States | 300 | 99% | English is the right answer here, not a fallback. 1 too short or too mixed to call. |
| United Kingdom | 151 | 99% | English is the right answer here, not a fallback. 1 too short or too mixed to call. |
| Italy | 493 | 99% | Fully measured in Italian. 4 too short or too mixed to call. |
| France | 4,506 | 93% | 4,233 answers in French, 224 in another language. 49 too short or too mixed to call. Every French answer since 2 July 2026 whose language could be read came back in French (4,123). |
| Germany | 149 | 77% | 115 answers in German, 32 in another language. 2 too short or too mixed to call. |
| Switzerland | 543 | 75% | 412 answers in German/French/Italian, 127 in another language. 4 too short or too mixed to call. Switzerland has three national languages, so we refuse to guess one: a Swiss prompt set has to name its language or it does not run. |
| Poland | 551 | 71% | 396 answers in Polish, 147 in another language. 8 too short or too mixed to call. |
| Portugal | 733 | 71% | 524 answers in Portuguese, 205 in another language. 4 too short or too mixed to call. The weakest market we have. |
Across all markets: 7,111 of 7,919 market-tagged answers (89%) came back in a language of their market. Language was detected on 22,263 of our 22,454 answers; the remaining 191 were too short or too mixed to call, and are recorded as unknown rather than assigned a language.
No AI-visibility tool covers everything, ours included. These are the gaps we know about. We'd rather you learn them here than discover them after signing.
| Surface | Status | Why |
|---|---|---|
| Meta AI | Not measured | Very large consumer reach, no API that permits measurement. This is a genuine blind spot for the whole category, not just for us. |
| Google AI Overviews / AI Mode | Read for some brands, not all | Distinct surfaces from the Gemini API, with different retrieval; we don't treat a Gemini API answer as a proxy for either. We read them from Google's results page through a data provider, not through a model's API, and only for the brands on that plan. Each is shown as its own engine, and both are also counted in the headline Mention Rate. A page can show no AI Overview at all. On the per-engine panel the rate counts it as a search that did not name you, as the headline does; the caption beside it says how often the Overview appeared and how often you were named when it did. |
| Perplexity, Grok, Mistral, DeepSeek, Llama | Built, piloted once, not tracked | The connectors work and we ran them once, on 18 August 2026: 81 answers from Perplexity and 20 each from the others. That is a single day and it has not been repeated, so nothing here is a trend. Those answers were collected for one brand and they do count in that brand's figures. When one starts running on a cadence it will be added to the engine table in section 02, with its model identifier and its count, before it is claimed anywhere else. |
| Microsoft Copilot | Never probed | No connector, no pilot, no stored answers. We have nothing to say about it. |
| Consumer apps vs APIs | Known divergence | We measure the chat engines through their APIs, and Google AI Overviews and AI Mode from Google's results page. The consumer apps add system prompts, memory and live retrieval, so their answers can differ from ours. We don't claim otherwise. |
| Model drift | Uncontrolled | Providers update models without notice. A change in your numbers can be a change in the model rather than a change in your visibility, which is why we record the exact model identifier on every stored answer. |
Run the same brand through two AI-visibility tools and you will get two different answers. That isn't one of them lying. It's four things stacking up:
If our number is lower than another tool's, the useful question is which prompt set each of us asked, and how many times. We'll show you ours.
The rules behind our recommendations are versioned and graded by evidence strength, from peer-reviewed research down to unverified industry folklore. When a rule is corrected or retired, the change is recorded with its date and its source rather than quietly edited — including the cases where we were previously wrong.
That ledger also runs in reverse: we keep an explicit list of widely-sold tactics the evidence does not support, so a recommendation can be refused as well as made.