Mention rate
Share of tracked prompts where the answer names your brand.
AI search visibility
AI visibility tracking and audits
Find out how often ChatGPT, Perplexity, Gemini, Google AI Overviews and AI Mode mention your brand, which pages they cite, and which competitors they recommend instead. Then track it every month on the same prompts.
Why track it
Your analytics only sees the clicks.
People often read an AI answer without clicking a link, so your analytics shows a fraction of what AI tells them about you.
Tracking the answers shows whether you're in them, how you're described, and who gets recommended instead.
01
Are you in the answer?
Mention rate and citation rate across a fixed set of buyer prompts.
02
Who's there instead?
Share of voice against the competitors you choose, by engine and by prompt group.
03
Where do the answers come from?
The domains and pages each engine cites, so you know which sources to pursue.
04
Is it accurate?
We log wrong prices, missing features and outdated descriptions, then trace the source and work to fix it.
What we track
Eight measures, every month.
01
02
Citation rate
Share of prompts where the engine links to your site as a source.
03
Share of voice
Your mentions against those of named competitors, by engine.
04
Answer position
Whether you're the first brand named or one of many.
05
Cited domains
The third-party sites each engine leans on for your category.
06
Accuracy and sentiment
How engines describe your product, pricing and category, and whether it's right.
07
Google rankings
Positions for the same topics, so you can compare search and AI.
08
AI referral traffic
Visits and conversions from AI assistants in GA4.
Our method
How we build and run the prompt set.
We write prompts from your keyword data, sales-call notes, support tickets and the words your customers use, then group them by buying stage: category questions, comparisons, problem-led questions and brand questions.
The set stays fixed month to month so the numbers are comparable. We add or retire prompts at each 90-day review and log every change.
01
Rates, not screenshots
AI answers change between runs. We report rates across the full prompt set and keep the raw answers, so one lucky or unlucky screenshot can't skew the picture.
02
Mentions and citations kept apart
Named in the answer and linked as a source are different results with different fixes, so we count them separately.
03
Settings recorded
Location and engine settings affect answers. We record the settings for each run so month-on-month comparisons stay fair.
04
Google alongside AI
We track Google rankings for the same topics, which shows whether AI visibility is following your search performance or lagging behind it.
Two ways to buy it
A one-off audit, or tracking every month.
Project
AI visibility audit
A one-off baseline across five engines with a competitor benchmark, a map of cited sources, a log of inaccurate answers and a prioritized fix list.
- Prompt set built for your category
- Share of voice against three competitors
- Fix list ranked by impact
Fixed feeQuoted after a scoping call
Monthly
Ongoing tracking
The same prompt set every month, with movement by engine, share of voice, cited sources and AI referral traffic, plus what we're doing about it.
- Monthly report and raw answers
- New and lost citations
- Included in GEO and SEO + GEO retainers
Included in retainersAlso available on its own
The monthly report
What lands in your inbox.
A report written for the people who make decisions, with the raw data attached for anyone who wants to check our work.
- Mention and citation rates by engine
- Share of voice against named competitors
- Prompt-level detail with raw answers
- New and lost citations
- Top cited domains in your category
- Accuracy issues and fixes
- AI referral traffic and conversions
- Priorities for next month
Common questions
Tracking questions.
ChatGPT, Perplexity, Gemini, Google AI Overviews and Google AI Mode. We also track Google rankings for the same topics, so you can see how the two relate.
Enough to cover your main categories, use cases, comparisons and competitors, usually somewhere between a few dozen and a few hundred prompts. We size the set in the proposal.
It's directional. Answers vary between runs, users and locations, so a single result means little. Rates across a fixed prompt set, tracked over months, show real movement.
Yes. We keep the raw answers for every tracked prompt, so you can read exactly what each engine said about you and your competitors.
Yes. We set up AI referral tracking in GA4 and connect it to your conversion events. It undercounts AI's influence, because many readers never click a link, which is why we track the answers too.
Yes, in every GEO and SEO + GEO retainer. You can also buy a one-off AI visibility audit or standalone monthly tracking.
Related services
Often paired with this service.
Get a proposal
See where AI puts you today.
Tell us your category and the competitors you care about. We reply within one business day to book a call, then send a proposal for an audit or monthly tracking.