How we measure Ai visibility.
Ai answers change from one run to the next, so a single screenshot proves nothing. Here is exactly how we test, score and track whether Ai engines mention, cite and recommend your business.
- Free, no obligation
- Four quick fields
- Plain-English findings
Four things in every answer
Mentions
Is your business named in the answer at all, and how prominently?
Recommendations
Are you presented as a top pick, or just listed among options?
Citations
Is one of your pages linked as a source, even when you aren’t named?
Who’s named instead
Which competitors appear, and which sources the engine relied on.
Every major Ai engine, side by side
The engines buyers actually use to research and shortlist businesses, plus Google Search for the same questions.
- Core enginesChatGPT, Gemini, Perplexity, Claude, Copilot, Grok and Google Ai Overviews, tracked as standard.
- Added where relevantMeta Ai, DeepSeek and Google Ai Mode, when your audience uses them.
- Search baselineGoogle Search results for the same questions, so Ai and traditional visibility can be compared.
Four steps, repeated every cycle
-
Build the question set
We agree a fixed set of buyer questions with you, from first research to “who should I hire”, using your sales calls, search data and competitors.
-
Run every engine, repeatedly
Each question runs in each engine more than once per cycle, in clean sessions with no chat history and with location set for local questions, because answers vary.
-
Score and map sources
Each answer is scored and every cited source is logged, showing which sites the engines trust in your market.
-
Re-test and report
The same questions are re-run on schedule, and the results land in your dashboard with a plain-English summary.
From answers to a score you can track
Each answer earns points based on how you appear in it. Your Ai visibility score is the points you earned divided by the points available, shown per engine and overall.
- Share of voiceHow often you’re named compared with the competitors who appear for the same questions.
- Source gapsSites cited for your questions where you aren’t present, ranked by how often engines use them.
| Result in an answer | What it means | Points |
|---|---|---|
| Recommended | Named as a top pick or first choice for the question | 3 |
| Mentioned | Named as one of the options, without being the lead pick | 2 |
| Cited only | Your page is a source, but your business isn’t named | 1 |
| Missing | Neither named nor cited | 0 |
Honest about what can be measured
No guarantees
No one controls what an Ai engine says. We improve the evidence and measure the effect.
Not every user sees the same thing
Personal history, location and wording change answers. Clean, consistent tests show the trend.
Models change
Engines update often. We note model and date with every run so changes can be explained.
Correlation isn’t causation
When a score moves, we say what changed and what we think drove it, and how confident we are.
Get your baseline
The free Ai visibility audit runs a starter set of your buyer questions through the major engines using this method, and shows who gets named, which sources are cited and what to fix first.
- Engine-by-engine checkMentions and citations across every major Ai engine
- Competitor comparisonWho gets named for your buyer questions
- Trusted-source mapReview sites, directories and publications Ai leans on
- Prioritized quick winsSite, schema, reviews and mention fixes, ranked
Request your free audit
Free and no obligation. We’ll email your findings in plain English.
Which Ai engines do you track?
ChatGPT, Gemini, Perplexity, Claude, Copilot, Grok and Google Ai Overviews as standard, plus Meta Ai, DeepSeek and Google Ai Mode where your buyers use them. We track Google Search results for the same questions so you can compare.
Why do you run each question more than once?
Ai engines don’t give identical answers every time. Repeating each question and reporting how often you appear gives a far more reliable picture than a single screenshot.
How often is visibility re-tested?
Monthly on every plan. Starter gets a monthly snapshot; Standard, Pro and Max track 25, 50 or 100 buyer prompts with competitor benchmarks.
Can you guarantee my score will go up?
No one can honestly guarantee what an Ai engine will say. We improve the signals engines rely on and report the results transparently, including when something doesn’t move.
Do you test logged in or logged out?
We use clean sessions with no chat history or memory, and set the location for local questions, so results reflect what a new buyer would see rather than a personalized answer.
Your next customer is asking Ai right now. Make sure it says your name.
Start with a free Ai visibility audit, or pick a plan and we’ll get to work.