Med Spa AI Recommendations: Why Booking Never Starts

Most med spa owners treat “best med spa near me” inside ChatGPT, Gemini, or Perplexity as a website problem. It isn’t. The assistant builds a short list from your Google Business Profile, your reviews, and what other sites say about you, then it stops. Your homepage is not the ballot. That is the same split I see in AI receptionist vs Google Business Agent: Google helps discovery, and the owned intake layer still has to catch the patient.

Short answer: Med spa AI recommendations come from your Google Business Profile, the words in recent reviews, and third-party mentions, not your website copy. SOCi’s 2026 index measured chain locations: ChatGPT recommended 1.2%, Gemini 11.0%, and Perplexity 7.4%. A one-location spa is not that statistic. Fix the profile so the model has accurate data, then book the patient who calls. An owned AI receptionist that does that booking is $8,000 once.

Why isn’t my med spa showing up when patients ask AI?#

The assistant skips you when your Google Business Profile, reviews, or listings are thin, wrong, or stale. ChatGPT, Gemini, and Perplexity build local names from that data, not your homepage. SOCi’s 2026 index measured chain locations, so the 1.2% ChatGPT rate is not one spa’s odds.

SOCi’s 2026 Local Visibility Index covered 2,751 multi-location brands and about 350,000 locations. A wrong category or stale reviews still gets a one-location spa skipped.

It also isn’t reading star ratings the way you’d guess. These models summarize what reviews say. If recent reviews mention “Botox,” “hydrafacial,” “no wait,” or “great with nervous first-timers,” the assistant can repeat those phrases when a patient asks for exactly that. If your reviews are thin or three years old, there’s nothing to summarize.

Here is how often those platforms recommended a studied location at all. The rates are from SOCi’s 2026 Local Visibility Index PDF, and Search Engine Land’s write-up of the same report:

Platform	How it sources local recs	Locations recommended	Profile accuracy
ChatGPT	Profile, reviews, and web mentions	1.2%	68.3%
Perplexity	Web index, reviews, and citations	7.4%	68.0%
Gemini	Grounded in Google Maps	11.0%	100%
Google local 3-pack	Maps pack, for comparison	35.9% appeared	—

Gemini recommends roughly ten times as often as ChatGPT in that study because it reads Google Maps. The PDF puts contact-data accuracy near 68% on ChatGPT and Perplexity and at 100% on Gemini. Locations ChatGPT recommended averaged 4.3 stars. SOCi’s own write-up also says single-location businesses often have cleaner local signals than a chain. Use the inputs. Do not treat 1.2% as your personal chance.

What does AI booking for a med spa actually do?#

AI booking for a med spa writes a real slot into GlossGenius, Boulevard, or Vagaro. The receptionist answers, asks which consult they want, and offers open times. It does not recommend a treatment. Clinical questions come to you by text. That owned desk is $8,000 once.

An AI Receptionist is that desk. Pricing exceptions and unhappy patients come to you by text with the note attached. A chatbot that only promises a callback is not booking.

Renting is a different bill. On Smith.ai’s AI receptionist pricing, checked October 2026, the Pro plan is $150 a month at $2.00 a call for 75 calls. The Enterprise plan is $500 a month at $1.67 a call for 300 calls. Overages are published, it is month to month, and canceling takes 30 days’ notice. Their page also lists AI scheduling against Calendly and similar tools. That is their stack, not your GlossGenius book.

Option	Published price	What you are buying
AI Receptionist, owned	$8,000 once	Your number, your script, booking into the tool you already run
Smith.ai AI, Pro	$150/month, $2.00/call, 75 calls	Their month-to-month receptionist. $1,800 if you keep it a year
Smith.ai AI, Enterprise	$500/month, $1.67/call, 300 calls	They build and run it. $6,000 if you keep it a year

The $150 plan is cheaper than $8,000 for years if 75 calls a month is really your volume. $8,000 against the $500 plan is about 16 months. I would not sell you the owned desk to “beat” a $150 subscription on price. I sell it when the phone has to book your calendar and you are tired of paying per call for a script you do not control.

The map I deploy:

• Trigger: A patient who found you through an assistant, Google, or a referral calls or texts, often after 6pm.
• AI action: The receptionist qualifies the request (“Botox consult, new patient, this week”), quotes only the prices you approved, and offers real slots.
• System of record: The appointment and a short note land in the booking tool and calendar. Nothing lives only in a chat thread.
• Human escalation: Clinical questions, a price exception, or an upset patient get flagged to you with the context attached.

If the leak is the form or the CRM note and the phone is already answered, that is Workflow Automation at $3,500 once, not the receptionist. Phone-first spas do not need both on day one.

What actually decides which med spas get named#

Four inputs decide the short list: your Google Business Profile, review volume and recency, identical name-address-phone across the web, and third-party mentions. None of them live only on your homepage, which is why a redesign does not get you named.

• Google Business Profile. Primary category “Medical spa,” not “Spa.” Every service listed by name. Hours accurate. Photos current.
• Reviews. Volume and recency both matter, and the words matter more than the star average. Ask patients to name the treatment. SOCi’s recommended ChatGPT locations averaged 4.3 stars, which is a floor, not a tactic.
• Consistency. Name, address, and phone identical on Google, Yelp, your site, and aesthetics directories. Conflicting data is how you become one of the profiles those assistants get wrong about a third of the time.
• Third-party mentions. A “best med spas in [city]” article, local press, or an industry directory. Those are the citations Perplexity and ChatGPT can point at.

If you only do one thing this quarter, make it the profile and reviews. The same plumbing decides which job to automate first. I wrote that up in picking the first med spa workflow.

The part nobody fixes: capturing the patient AI sends you#

Getting named is half the sale. The other half is answering when that patient calls, texts, or submits a form, often after hours. If the contact hits voicemail, they book the next name on the list and your profile work paid a competitor.

AI discovery does not book anyone. It only moves a patient to reach out, often within minutes.

For a three-room med spa I worked with, the booking software was fine and about 40% of inbound calls still arrived after 6pm with nobody to answer. The fix was not another ad. It was a receptionist that books while the owner is in a treatment room. If you want the calculator on what missed calls cost in your numbers, use the missed-call cost calculator.

Visibility without capture is just a faster way to feed your competitors. That is the same leak as any AI lead generation project: close it before you pour more demand into the bucket.

What I’d fix first#

Fix the Google Business Profile and fresh treatment-specific reviews this week, then close the after-hours gap. Set the category to Medical spa, list every service, and post current photos. Capture is next, or the patients who do get your name still bounce.

Ask satisfied patients for a review that names the treatment. That is the language the assistants can repeat.

Then close the after-hours gap so the patients who do get your name do not bounce. The shape I would build is the med spa after-hours booking setup.

When chasing AI visibility isn’t the right move yet#

Skip the AI-search project if the phone, the booking tool, or the Google profile is already broken. Unanswered daytime calls, an unclaimed profile, and a calendar that is not set up all beat chasing which assistant names you. Assistants send a trickle, not a flood.

The 1.2% figure is a chain-location rate, not a reason to reorganize the spa.

Also wait if you cannot staff more volume. Extra bookings you cannot deliver become no-shows, rushed rooms, and bad reviews, which quietly lowers the review quality those models summarize. Get room utilization and front-desk coverage solid first.

Where to start#

Clean the Google profile, then make every call and text book into the calendar you already trust. If patients are finding you through an assistant, close the capture gap. Rebuilding the website does not change what ChatGPT, Gemini, or Perplexity cite.

Want me to look at your setup? Send a free audit and I’ll reply with your AI replacement map within 24 hours, no call required. If you’d rather talk it through, book a 20-minute call.