ChatGPT-referred clients may convert better than Google clients at a med spa, but no med spa study proves it. Cross-industry data from Microsoft Clarity, Adobe, and Ahrefs shows AI referrals converting above search per visit, on volumes under 1 percent of traffic. None of it measures booked appointments, show rates, or memberships.
AEO, GEO, AI visibility, and AI SEO are four labels for one job: getting a business named and cited when someone asks ChatGPT, Perplexity, Gemini, or Google AI Overviews for a recommendation. This post uses AEO. The [plastic surgery version of this question](/blog/chatgpt-referred-patients-vs-google-search-conversion-plastic-surgery) judges conversion on consult-to-surgery rate for a one-time procedure. A med spa is judged on something longer: whether an AI-referred client shows up, comes back in three to four months for the next injectable visit, and joins a membership. The studies below are a different set from the ones in that post.
- No independent study found while writing this post compares ChatGPT-referred med spa clients with Google search clients. The closest evidence comes from publisher, retail, and software sites.
- Microsoft Clarity's 2025 study of publisher and news sites found AI referrals were under 1 percent of traffic but converted at about 3 times the rate of search and social, and Adobe found AI referrals to U.S. retail sites converted 31 percent more than non-AI traffic in the 2025 holiday season.
- For a med spa the conversion that matters is not the first booking but the client who shows, rebooks, and joins a membership worth $120 to $280 a month. No published study measures AI-referred clients on show rate or membership conversion, so a spa has to measure it from its own booking data.
What does the 2025-26 conversion data actually show?
The 2025-26 data shows AI referrals converting above other sources per visit, across three independent datasets, none of them from healthcare. Microsoft Clarity, Adobe Analytics, and Ahrefs each measured a different conversion event on a different kind of site, so the direction agrees while the size of the gap does not.
Microsoft Clarity analyzed one month of data from publisher and news sites and reported that AI referrals made up less than 1 percent of traffic while converting at about 3 times the rate of search and social. Trade coverage of the same study reported subscription conversion at 1.34 percent for AI referrals against 0.55 percent for search. Adobe Analytics reported that from November 1 to December 31, 2025, AI referrals to U.S. retail sites converted 31 percent more than non-AI traffic, after converting 9 percent less in the 2024 holiday season. Ahrefs reported that on its own site in June 2025, AI search sent 0.5 percent of visitors and 12.1 percent of signups.
| Source | Sites measured | Conversion event | AI referral result |
|---|---|---|---|
| Microsoft Clarity, 2025 | Publisher and news sites, one month | Subscriptions and sign-ups | About 3x search and social; 1.34% vs 0.55% on subscriptions |
| Adobe Analytics, holiday 2025 | U.S. retail sites | Purchase | 31% higher than non-AI traffic |
| Adobe Analytics, holiday 2024 | U.S. retail sites | Purchase | 9% lower than non-AI traffic |
| Ahrefs, June 2025 | One software site (its own) | Account signup | 0.5% of visitors, 12.1% of signups |
Why might an AI-referred med spa client convert better?
An AI-referred med spa client may convert better because the shortlisting already happened inside the chat. A client who asks ChatGPT which nearby spas offer Daxxify, what lip filler costs, or which injectors have the most experience arrives having compared options, while a Google searcher often opens four tabs and starts comparing on arrival.
That is a hypothesis, not a finding. The Adobe numbers are a caution: the same AI channel converted below other traffic one holiday season and above it the next, which suggests user behavior and answer quality are still shifting. A spa should expect its own rate to move as well.
Why might it convert worse?
An AI-referred med spa client may convert worse when the answer that sent the client was wrong. AI answers sometimes quote outdated prices, list a treatment the spa has dropped, or name an injector who has left, and a client who books on that information is more likely to cancel or leave unhappy after the first visit.
Price-led prompts carry a second risk. A client who asked for the cheapest Botox in town converts on price and may not rebook when a competitor runs a lower unit special. Whether AI-referred clients skew toward price shoppers or toward researched buyers is unmeasured, and the answer probably differs by spa and by prompt.
Which conversion should a med spa measure?
A med spa should measure four conversions by source, not one: booking, show, rebooking, and membership. A source that books well but shows poorly, or brings first visits that never return, costs more than its booking rate suggests, because a med spa earns most of its revenue from repeat visits rather than the first appointment.
- 1. Booking rate: inquiries or site visits that become a booked appointment
- 2. Show rate: booked appointments that arrive, measured over at least 60 days
- 3. Rebook rate: first-visit clients who book a second visit within 120 days, which covers a typical neurotoxin cycle
- 4. Membership rate: first-visit clients who join a membership within 90 days, the metric tied to $120 to $280 a month in member LTV
Does volume matter more than conversion rate?
Volume matters more for most med spas today, because AI referrals are still a small slice of traffic in every published dataset. A higher conversion rate on a small base adds a few clients a month, while the same improvement on Google search or Meta changes the month.
For example, a spa booking 50 first appointments a month from Google search, with 20 percent joining a membership, adds 10 members. If AI answers send 6 bookings and 35 percent join, that adds about 2 members. At $120 to $280 a month, those 2 members are $240 to $560 in monthly recurring revenue, which is worth having but is not a reason to move budget. The figures are illustrative, not measured rates.
What has ClinicAds seen in its own data?
ClinicAds has a labelled observation from its own website rather than from client spas. As of August 2026, chatgpt.com was the largest referring site to clinicads.com in Vercel Web Analytics, and the first inbound sales call ClinicAds traced to an AI referral came from a chatgpt.com visit. That is one agency site with small numbers, not a benchmark, and no client data is published here.
The more useful observation is about attribution. A client who hears about a spa in ChatGPT often searches the spa's name before booking, so analytics records the visit as branded search or direct. Referrer data alone undercounts AI-referred clients, which is why a self-reported source question matters.
How should a med spa set this up?
A med spa should capture source at booking, carry it into the client record, and compare cohorts after 90 days. Most booking platforms used by spas, including Boulevard, Zenoti, and Vagaro, support a custom intake field, and that field is the only way to catch clients whose AI referral shows up as a branded search.
ClinicAds plans med spa paid media at $28 to $80 per booked appointment and 4 to 8x return on $3,000 to $8,000 a month. Those are agency averages, not guarantees, and they are the baseline an AI-referred cohort should be compared against. The [med spa marketing program](/medspa) runs AEO inside the same reporting, not as a separate service.
- Add a required booking question, how did you hear about us, with ChatGPT or another AI assistant as a listed answer
- Tag chatgpt.com, perplexity.ai, gemini.google.com, and copilot.microsoft.com referrals in GA4
- Keep a monthly log of whether the spa is named for its top 15 treatment and membership prompts
- Wait for at least 30 first visits per source before comparing membership rates
Do ChatGPT visitors convert better than Google visitors?
In the published 2025 data, usually yes per visit. Microsoft Clarity found AI referrals converting at about 3 times search and social on publisher sites, and Adobe found a 31 percent edge in 2025 holiday retail. No study covers med spas.
Is there a study of ChatGPT-referred med spa clients?
None was found while writing this post. A med spa can build its own comparison with a booking-source question and 90-day cohort tracking of show, rebook, and membership rates.
Why do AI-referred clients show up as direct or branded traffic?
Many clients read a spa's name in ChatGPT, then search that name on Google or type the website address. Analytics records the last step, so the AI referral disappears unless the booking form asks.
Should a med spa cut Google spend if AI clients convert better?
No. AI referrals are a small share of traffic in every published dataset, so a better rate adds few clients. Keep paid and search budgets on their own returns and treat AEO as an added source.