Oracite Research · June 2026
We audited 6 Silicon Valley med spas. Most are invisible in AI search.
We asked ChatGPT, Perplexity, Gemini, and Google AI Mode the questions patients actually ask — and recorded 216 question–answer pairs. The typical clinic was named in only about a third of answers, and one with 587 reviews showed up just 14% of the time.1
original data · six clinics · 216 question–answer pairs · every run sealed with a tamper-evident receipt
What we found
Patients increasingly pick a provider by asking AI — “best Botox near me,” “is laser worth it,” “safest place for fillers.” We asked those exact questions on every major engine, for six clinics, and recorded who got named and cited. Across 216 question-answer pairs, a clinic was mentioned only about 32% of the time. The rest went to someone else.
And the gap was stark. One clinic that had quietly done the right things showed up almost everywhere — while a clinic with 587 reviews was named in only 14% of answers. Reputation and ad spend didn’t decide it. Doing the things AI rewards did.
Google’s AI is where it’s decided
Google AI Mode now has over a billion monthly users2 and shows an AI answer on the majority of “how much / is it worth it” local searches. It’s the biggest audience — and even there, the clinics we tested were named less than half the time, while it routinely listed a dozen competitors per question.
The backdrop: search is going zero-click
Share of U.S. Google searches that end without a click
Source: SparkToro zero-click search studies — clickstream panels from Jumpshot (2016, 2019), Datos (2024) and Similarweb (2026); cross-year series compiled by a16z. Primary: SparkToro, 2026. Chart by Oracite.
Who AI recommends instead
Across the audits, a handful of names kept winning the answers:
The lesson isn’t “get more reviews.” It’s that AI builds its answer from sources it trusts — and most clinics have given it nothing of theirs to trust.
How we measured it
For each clinic we ran a fixed set of buying-intent questions across ChatGPT, Perplexity, Gemini, and Google AI Mode, captured every answer verbatim, and recorded whether the clinic was named and cited. Every run is logged with a SHA-256 receipt, so findings are reproducible and tamper-evident. It’s a small, transparent sample — six Silicon Valley clinics, 216 question–answer pairs, from our own audits — not a claim about the whole industry, but the pattern is consistent and rarely measured.
method: fixed prompt set · 4 engines · verbatim capture · sha-256 receipt per run · june 2026
Questions
How visible are aesthetic clinics in AI search? In our audit of six Silicon Valley clinics across 216 question–answer pairs, the typical clinic was named in only about a third of answers (~32%). One clinic with 587 reviews appeared in just 14%.
Why isn’t my med spa showing up in ChatGPT or Google AI? AI engines build answers from sources they trust — review platforms, directories, structured data, consistent entity info. Reviews and ad budget alone don’t get you named.
Can this be fixed? Yes. The clinic that scored well wasn’t bigger — it had done the things AI rewards. That’s exactly what Oracite does: audit where you stand, earn the citations, and prove the lift each month.
See where your clinic stands — free.
We’ll run your AI Visibility audit and show you, word for word, what AI tells your patients.
Sources
- Oracite audit, June 2026 — 6 Silicon Valley aesthetic clinics across 216 buying-intent question–answer pairs on ChatGPT, Perplexity, Gemini, and Google AI Mode; every run logged with a SHA-256 receipt. Methodology in “How we measured it” above.
- Google, “How people are using AI Mode,” Google I/O 2026 — AI Mode has surpassed 1 billion monthly active users. blog.google