How to rank your medspa on ChatGPT

1 · The premise

ChatGPT’s default consumer model now answers local recommendation questions with web tools enabled. A patient asking “best botox in Sunnyvale” gets a compressed answer naming three to five specific clinics, often with a short rationale and sometimes a citation link.

The patient does not see the candidates that did not make the list. If your clinic is one of them, the query effectively does not exist for you. This guide is about getting into the list. The deeper service-side view of ChatGPT optimization is the complement to this post.

2 · How ChatGPT decides who to recommend

ChatGPT’s ranking logic is unpublished. The observable inputs cluster into three families:

• Retrieval. Which sources does ChatGPT’s search layer surface for the query? Major directories, local press, the clinic’s own site, third-party review aggregators.
• Authority. Are those sources well-cited elsewhere? Is the clinic’s own URL internally consistent? Does structured data confirm what the prose claims?
• Specificity. Does the clinic’s presence match the query intent? A treatment-led query rewards clinics that name the treatment explicitly on a retrievable page.

The seven steps below address all three families. They are ordered roughly by leverage: the highest-impact fixes near the top.

3 · The 7-step playbook

1. Make your treatments and locations explicit on your site. One page per major procedure (botox, filler, microneedling, PRP, laser resurfacing, etc.) with the procedure name, the city, and a real price band in the on-page prose. ChatGPT’s retrieval layer cannot resolve a treatment-led query against a clinic whose own site does not name the treatment in retrievable text.
2. Get cited by third parties. A Google Business Profile is necessary but not sufficient. Add Yelp, RealSelf where relevant, one or two local press mentions, and any aesthetic-vertical directory ChatGPT is already pulling for your metro. ChatGPT downweights clinics with no corroborating sources outside their own site.
3. Implement LocalBusiness, Service, and FAQPage structured data. JSON-LD schema lets the retrieval layer confirm what the prose claims. LocalBusiness with the canonical name, address, and service area. Service entries for each procedure. FAQPage for common patient questions. Keep the schema consistent with the visible page content: contradictions get downweighted.
4. Build topical authority per treatment. Beyond the procedure landing page, add supporting content: short explainers, before-and-after gallery pages annotated with treatment names, comparison pages ("botox vs dysport") that name your clinic in context. This is slower than the technical fixes but compounds.
5. Get reviews that mention treatments by name. ChatGPT’s retrieval favors review corpora with semantic specificity. A review that says "best filler experience in Sunnyvale" carries more signal than a generic five-star with no text. Coach booking flow to surface a review request with a hint at what to mention.
6. Tighten name-address-phone consistency. Across Google Business Profile, Yelp, your site, and any directory you appear in, the clinic name, address, phone, and listed services should be identical. ChatGPT downweights entities whose sources disagree about basic facts.
7. Monitor monthly. ChatGPT updates continuously. A clinic that ranks in the top three this month may drift to fifth next month if a competitor ships fixes or earns new citations. Monthly re-scoring catches drift before it costs you bookings.

4 · What this doesn’t do

These seven steps move a clinic into ChatGPT’s consideration set. They do not guarantee position one on any specific query: ChatGPT’s ranking is stochastic at the margin, and competing clinics ship fixes too. The realistic target is: be one of the three to five names that get returned, consistently, across the patient-intent queries that drive consults in your metro.

They also will not move Gemini, Claude, Google AI Overviews, or Perplexity by themselves. Those surfaces share some signals with ChatGPT (structured data, citation consistency) but weight them differently. A real AEO program covers all five. See the methodology for the cross-surface picture.

5 · The shortcut

The seven steps are doable in-house. Most clinics do not have the bandwidth to ship them, then measure the result, then re-prioritize monthly. That is the case for the audit: ranked fixes ordered by projected lift, re-scored every 30 days. If you want the structured version, the AI visibility audit is the right starting point. If you want to confirm there is a problem first, the 10-minute self-check answers that.