How it works
AI mention tracking is closer to a poll than a report. You build a fixed list of prompts that match how patients actually ask, run them on a schedule, and record what comes back.
- Build the prompt set. Use real patient language, not marketing language. "Best Botox near me in Austin." "Who does natural-looking lip filler in Toronto." "Is CoolSculpting worth the money."
- Pick the assistants. ChatGPT, Google AI Overviews, Gemini, Perplexity, Copilot. Each pulls from different sources and each has its own habits, so results rarely match.
- Run on a schedule. Same prompts, same wording, same locations, every month.
- Record the answers. Was your clinic named? Where in the answer did it land? Which page, profile, or article did the assistant cite to get there?
- Score it. Your mention rate is the share of prompts where you appear. Then compare that rate against the competitor median for the same set.
The fixed prompt set is the whole point. AI answers drift from run to run, so a single spot check tells you almost nothing. Asking the same questions the same way, month after month, turns that noise into a trend line you can act on. The citation log matters just as much as the score. It shows you which sources the assistants actually trust in your market, which is where the work goes next.
Why it matters for aesthetic clinics
Aesthetic treatments are high-cost and high-consideration. Patients research for weeks before they book, and a growing share of that research now starts with an assistant instead of a results page. They ask, they read one answer, and they never see the ten blue links.
That changes the math. A search results page gives you many chances to be found. An AI answer usually names only a handful of clinics, often three to five, and everyone else is simply absent. There is no page two to be on. You are named or you are invisible, and nothing in your analytics will tell you which.
This is the gap mention tracking closes. Without it, you are guessing about the single channel where your best-fit patients are forming their shortlist. With it, you know whether you are on that shortlist, how you compare to the clinics down the road, and which sources are putting them there instead of you.
AI mention tracking vs AI share of voice
The two get used interchangeably. They answer different questions.
| AI mention tracking | AI share of voice | |
|---|---|---|
| What it counts | Whether you appear at all | How much of the answer space you own against rivals |
| Output | A mention rate across your prompt set | A percentage split across all named brands |
| The question it answers | Are we in the room? | How loud are we compared to everyone else? |
| Best used for | Early diagnosis and spotting blind spots | Ongoing competitive benchmarking |
Start with mention tracking. Share of voice only becomes useful once you appear often enough to have a share worth measuring.
The Ownerized take
Mention tracking is a thermometer, not a treatment. It tells you that you are invisible in AI answers, which is worth knowing, but it does not make you visible. The clinics that move are the ones that treat the citation log as the real output: it names the exact directories, review platforms, and pages the assistants already trust, and that list becomes the work order. We track mentions as the scoreboard, then fix the sources feeding it, which is how the AI Growth System is built.
Common mistakes
- Checking once and calling it data. One run is a coin flip. A mention rate only means something across a stable prompt set over time.
- Writing prompts a marketer would use. "Premium aesthetic solutions provider" is not a phrase any patient has typed. Use their words.
- Ignoring location. Assistants answer differently by city. A prompt without a place is a prompt without an answer you can use.
- Tracking the score and skipping the citations. The score tells you where you stand. The citations tell you what to do about it.
- Chasing every assistant equally. Find out which one your patients actually use before you spend a month optimizing for the others.
- Comparing yourself to a national brand. Your benchmark is the clinics competing for the same appointment slots, not a chain with a press team.
Frequently asked questions
How often should we run AI mention tracking?
Monthly is enough for most clinics. AI answers shift week to week, so weekly checks mostly capture noise and burn staff time. Run the same prompt set on the same schedule, and compare month over month. If you launch a large content or PR push, add one extra check four to six weeks after.
How many prompts do we need to track?
Enough to cover your money treatments across your service area, usually a few dozen. Include treatment questions, near-me questions, comparison questions, and safety questions. Too few prompts and one lucky answer swings your whole score. Too many and you spend hours reading answers no real patient would ever ask.
Can we pay to appear in AI answers?
Not directly, the way you buy a Google ad. Assistants build answers from sources they already trust: your site, directories, review platforms, press, and forums. The only reliable lever is being the clearly correct answer in those places. Some assistants are testing ad formats, but that is separate from being named organically.
Is AI mention tracking different from tracking Google rankings?
Yes. Rankings tell you where a page sits in a list of links. Mention tracking tells you whether an assistant names your clinic inside a written answer, and which sources it leaned on to do it. A clinic can rank well and still go unmentioned, which is the gap worth finding.
What is a good mention rate for a med spa?
There is no universal number, and anyone quoting one is guessing. The only useful benchmark is your local competitor median on the same prompt set. If three rivals get named on most prompts and you get named on a few, that gap is your target, not an abstract percentage.