Home / GEO Services / AI Visibility Tracking
Tracking
AI Visibility Tracking
Every month, a fixed set of buying prompts is re-tested across five AI platforms and each answer is scored as absent, source, named or recommended, so progress is measured in what buyers actually see.
The four rungs
What each rung means
| Rung | What happens in the answer | What usually causes it | Typical fix |
|---|---|---|---|
| 00 Absent | No page from the brand is retrieved | Blocked crawlers, missing indexation, no page on the topic | Access, indexation, new page |
| 01 Source | A page is cited, but the brand isn't named | The lifted passage never mentions the brand or product | Passage rewrite, naming |
| 02 Named | Listed with competitors | Weak comparison content, thin third-party coverage | Comparison pages, review and directory presence |
| 03 Recommended | The brand is the answer | Strong answers, consistent entities, trusted sources | Maintain and monitor |
Diagnosis
Four reasons a brand gets stuck
Access
The platform can't reach or read the page.
Retrieval
The page isn't indexed or isn't relevant enough to be pulled in.
Attribution
The page is used, but the brand isn't credited.
Preference
Other sources are trusted more for the recommendation.
Metrics
Three shares, reported monthly
Citation share
The share of prompts where at least one of the brand's URLs is cited.
Mention share
The share of prompts where the brand is named in the answer text.
Recommendation share
The share of prompts where the brand is the first or only recommendation.
FAQ
Method questions
What is AI visibility?
AI visibility is how often, and how favourably, a brand appears in answers from AI platforms such as ChatGPT, AI Overviews, Gemini, Perplexity and Copilot: whether it is cited, named or recommended.
How is AI visibility tracked?
A fixed set of buying prompts is re-run every month across each AI platform in clean sessions. Every answer is classified as absent, source, named or recommended, and the brands named and URLs cited are recorded, so changes show up as shifts in share rather than one-off results.
How is brand visibility tracked in Google AI Mode?
AI Mode is tested with the same prompt set as the other platforms, in a clean session set to the target market. Because AI Mode splits questions into many sub-searches, the cited URLs are logged to see which pages answer which parts of the question.
How is brand visibility measured in AI answers?
With three shares across the prompt set: citation share (a brand URL is cited), mention share (the brand is named) and recommendation share (the brand is the first or only recommendation). AI referral traffic and leads are reported alongside.
Why track AI brand visibility?
Buying decisions increasingly happen inside AI answers, where rank trackers can't see. Tracking shows whether the brand is on the shortlist, which competitors are winning, and whether changes to the site are working.
Can AI answers be measured reliably when they change between runs?
Single answers vary. Testing a fixed prompt set in clean sessions and tracking shares across the whole set gives a stable trend month to month.
How is the prompt set built?
From Search Console queries, sales and support questions, and competitor comparisons, grouped into definitions, shortlists, comparisons, how-to, regional and brand prompts. The core set stays fixed so trends are meaningful.
See where the brand stands in AI answers.
The AI Visibility Audit gives every buying prompt a visibility level and explains why.