Why ChatGPT Doesn't Recommend Your Med Spa (And What Actually Fixes It)
Someone opens ChatGPT and asks “what's the best med spa near me for Botox?” The model doesn't roll dice. It runs a series of checks against every candidate business, and it only names the ones that pass. Here's what those checks actually are.
ChatGPT composes an answer — it doesn't search a list
When ChatGPT answers a local question, it draws on its training, a live web/search tool, and the structured, authoritative sources it trusts: business profiles, well-known directories, review platforms, and entity-rich web pages. It tends to name a small number of businesses and explain why each one fits. If your business never clears the bar on those sources, it never enters the shortlist — regardless of how good your actual service is.
The five checks a candidate business has to pass
- Does this business clearly exist as a real entity? The model wants a complete, claimed, authoritative profile — the Google Business Profile is the anchor — with category, services, hours, location, and photos. A thin or unclaimed profile reads as low-confidence and gets skipped.
- Are its core facts consistent everywhere? Name, address, and phone (NAP) need to match byte-for-byte across your website, your Google Business Profile, and every directory. A mismatched suite number or an old phone number on one listing is enough to introduce doubt — and doubt loses to a competitor whose facts line up cleanly.
- Is it present in the sources the model trusts for this vertical?For med spas that means Yelp, RealSelf, Healthgrades, and Apple Business Connect, plus “best med spa in [city]” roundups and local press the model has indexed.
- Is it structured so a machine can extract it? LocalBusiness/MedicalBusiness schema, per-treatment pages with service-level markup, and FAQ markup let the model pull clean facts instead of guessing from prose.
- Is it trusted and active? Recent, responded-to Google and Yelp reviews signal a real, well-regarded, currently operating business. Review velocity and owner responses matter more than a single high star count from years ago.
These aren't five equally-weighted boxes to check in whatever order is convenient. They're sequential. There's no point chasing a mention in a local magazine if your Google Business Profile is incomplete or your NAP is inconsistent — the model can't yet confidently identify your business at all, so a third-party mention won't rescue it.
The fix order that actually moves the needle
Work top-down. Foundational fixes first, because everything above them depends on the model being confident your business exists and is exactly who it says it is:
- Complete and optimize your Google Business Profile — categories, every service listed, current hours, recent photos.
- Fix NAP consistency and get listed (or correct existing listings) on the directories AI actually cites.
- Add LocalBusiness/MedicalBusiness schema and dedicated pages per treatment.
- Build review velocity — steady, recent, responded-to reviews, never bought or incentivized.
- Earn third-party mentions and “best of” listicles once the foundation is solid.
This is cumulative work, not a switch you flip. Profile completeness and NAP consistency can show up in AI answers within a few weekly re-scan cycles. Reviews and third-party mentions compound over weeks and months. The honest expectation is steady lift, not an overnight ranking jump.
Before guessing which of these five checks you're failing, run the same query a customer would — and see exactly who ChatGPT names instead of you.
Check where you stand right now→Ready?
See exactly why AI is (or isn't) naming you.
A free scan shows your citation share on ChatGPT and Google AI Overviews, plus the sources it's pulling from.