A good AI visibility score for a med spa is 35 to 60 percent coverage across two engines with a win rate above 5 percent, scored per treatment line rather than per clinic. Below 15 percent coverage the med spa is missing from the sources assistants retrieve. Above 60 percent with a win rate over 20 percent, the work changes from building position to holding it.
ClinicAds has already published the measurement protocol for plastic surgery practices, covering prompt set design, run counts, and the split between coverage and win rate. This post does not repeat that protocol. It answers the half of the question that post left open, for med spas specifically: what the resulting number should look like, why a med spa needs a third number a surgical practice does not, and why one clinic-level percentage reports less than it appears to.
- A med spa holding 35 to 60 percent coverage across two engines, with a win rate between 5 and 20 percent, is inside the retrieved source set and competing on ordering. Below 15 percent coverage the clinic is absent from the source set entirely.
- A med spa needs three numbers where a surgical practice needs two. The third is the answerable rate: the share of logged runs where the assistant names any clinic at all. Runs that name nobody are not losses and must be removed from the denominator before coverage is computed.
- Coverage scored per clinic hides the result that matters. A med spa can hold 60 percent coverage on neurotoxin queries and 0 percent on body contouring, and the blended number reports neither.
- There is no published med spa AI visibility benchmark. The only defensible comparisons are a clinic's own dated baseline and the competitors named inside the same logged runs.
What is a good AI visibility score for a med spa?
A good med spa AI visibility score is 35 to 60 percent coverage across at least two engines, with a win rate above 5 percent, measured on the clinic's lead treatment line. No public benchmark exists for this metric, because the number depends entirely on which prompts a clinic chose to run. A med spa that scores its own brand name will read 90 percent and learn nothing.
The two comparisons that hold up are a med spa's own dated baseline and the competitor names appearing inside the same logged runs. If three competing clinics are named in eight of ten runs and the med spa appears in two, that gap is the finding, and it is readable without any industry average at all.
- Answerable rate: the share of logged runs where the assistant names any clinic at all
- Coverage: the share of answerable runs naming this med spa anywhere in the answer
- Win rate: the share of answerable runs where this med spa is the first or sole recommendation
Why do med spa queries need a third number?
Med spa queries produce a high share of answers that name no business at all. An assistant asked how to choose an injector, or asked about a clinic problem in a patient's own words, frequently returns generic guidance with zero brands in it. Those runs are not losses for the med spa. Counting them in the denominator drags coverage down and makes a clinic look beaten in a category nobody won.
ClinicAds measured this pattern on its own agency category on 2026-08-13. Five of the logged Perplexity runs were med spa or aesthetic-category queries. Two of those five returned no brand names at all. The capture below is a method demonstration on the agency category rather than clinic data, and it is one dated observation, not a trend. It is included because the 40 percent unanswerable share is the part a med spa has to handle before any coverage figure means anything.
- 2 of 5 med spa category runs that day named no business at all, a 40 percent unanswerable share
- An unanswerable query is a first-mover surface, not a defeat: nothing is currently being retrieved to answer it
- Report the answerable rate alongside coverage, since a rising answerable rate means competitors are filling the gap
| Query axis | Any brand named | What came back | How the run scores |
|---|---|---|---|
| Category, best agency for med spas | Yes | Six competitors named, ClinicAds absent | Answerable, counts against coverage |
| Vocabulary gap, aesthetic and medical aesthetics clinics | Yes | Seven competitors named, ClinicAds absent | Answerable, counts against coverage |
| Flagship, AEO and AI search for med spas | Yes | Three named, two of them new entrants | Answerable, counts against coverage |
| Problem language, plenty of leads but nobody books | No | Generic advice, no business named | Unanswerable, removed from the denominator |
| Offer wedge, one-time install instead of a retainer | No | The model described, no business named | Unanswerable, removed from the denominator |
Should a med spa score coverage per treatment or per clinic?
Per treatment line. A med spa sells five or six distinct treatment categories, and an assistant retrieves a different source set for each one. Neurotoxin queries pull from manufacturer provider locators and booking platform menus. Body contouring queries pull from device manufacturer pages and metro cost roundups. Medical weight loss queries pull from an entirely separate set again, most of it built after 2024.
A blended clinic-level percentage averages those independent results into a number that describes no actual query. A med spa reading 30 percent coverage overall may be at 60 percent on neurotoxin and 0 percent on the body contouring line it just spent $90,000 on a device to offer. The blended figure conceals the only result that would have changed a decision.
- Score each treatment line separately: neurotoxin, filler, laser and skin, body contouring, medical weight loss, membership and facials
- Run 4 to 6 prompts per treatment line, 3 runs each, across 2 engines
- Report the lead treatment line first, since it carries the appointment volume the clinic is actually defending
- Track any newly launched service as its own line from launch day, so there is a baseline from before the marketing started
What coverage targets fit a med spa's stage?
Targets move with where a med spa already sits in the retrieved source set, not with revenue or clinic size. A clinic absent from booking platforms and manufacturer locators cannot reach 50 percent coverage through on-site work, because the sources an assistant reads for med spa answers are mostly not the clinic's own website. The four stages below describe what each coverage and win rate pair calls for next.
| Stage | Coverage | Win rate | The lever that moves it |
|---|---|---|---|
| Absent from the source set | 0-15% | 0% | Google Business Profile accuracy, booking platform menu with prices, manufacturer provider locator listings |
| Listed but rarely retrieved | 15-35% | 0-5% | Treatment-level review text and inclusion in metro best-of roundups for the lead treatment |
| Retrieved and competing on order | 35-60% | 5-20% | Per-treatment depth, published price ranges, and recency of the sources describing the clinic |
| Holds the metro on a treatment line | 60%+ | 20%+ | Weekly re-runs, competitor entry monitoring, and extending the same work to a second treatment line |
How much is med spa AI coverage worth in bookings?
The value of coverage for a med spa is set by membership rather than by a single appointment. ClinicAds works to a planning range of $28 to $80 per booked appointment on managed paid media at $3,000 to $8,000 per month, at 4 to 8x return on ad spend, with member value in the range of $120 to $280 per month. Those are agency averages, not guarantees. A member retained six months therefore carries $720 to $1,680 in value against the booking cost that started it.
An assistant recommendation carries no media cost, which is the argument for tracking the surface even while it delivers a fraction of paid search volume. The honest caveat is timing. Med spa coverage responds to third-party surfaces being updated, re-crawled, and then retrieved, so it moves over quarters. A clinic that starts logging today is buying a dated baseline, which is the cheapest and least reversible part of the program.
How often should a med spa re-run the check?
Weekly on the lead treatment line and monthly across every line. A med spa menu changes faster than a surgical practice menu, with new devices, seasonal offers, and membership tier edits, and each change alters what a booking platform and a manufacturer locator publish about the clinic. A monthly full run recomputes answerable rate, coverage, and win rate across all lines and is the figure that belongs in a quarterly review.
Sample size decides whether the number can be defended. ClinicAds holds a med spa result to 30 or more logged runs, two or more engines, and a span of at least 21 days before treating it as a position rather than a reading. Every run gets logged, including the ones naming no clinic at all, because the runs that are discarded are exactly the ones that made the number look good.
- Weekly: 4 to 6 prompts on the lead treatment line, 3 runs each, 2 engines, roughly 30 logged runs
- Monthly: every treatment line, recompute answerable rate, coverage, and win rate
- Within 14 days of any menu, price, or device change: re-run the affected treatment line
Is 40 percent coverage good or bad for a med spa?
It depends on the treatment line and on who else is in the log. A med spa at 40 percent coverage on neurotoxin queries, in a metro where the next clinic sits at 15 percent, holds a real advantage. The same 40 percent against two competitors at 70 percent is a deficit. ClinicAds reads the number against the competitor names inside the same runs rather than against an industry figure that does not exist.
Why remove unanswerable runs from the coverage denominator?
Because a run where the assistant names no clinic at all was not won by anyone. Leaving those runs in the denominator understates coverage and makes a med spa look beaten in a query no competitor took either. ClinicAds reports the answerable rate as its own number, since a query that names nobody today is a surface a med spa can take before anyone else does.
Can a med spa track AI visibility with a rank tracking tool?
Only the site-side half. Schema validity, crawl access for AI user agents, and answer-block structure are all verifiable from the clinic site and can run daily without a human. The engine half stays manual, because ChatGPT, Perplexity, and Google restrict anonymous automated querying and return degraded results that look like data and measure nothing.