How ChatGPT, Perplexity & Gemini Recommend Med Spas | Cornflower

Written by Nava Atkinson | Mar 30, 2026 7:35:49 PM

Open ChatGPT and type "best med spa for lip filler in [your city]." In under three seconds, you'll get a list of three to five names, a sentence about each, and sometimes a link. That list is now influencing real booking decisions — and the way it's generated is nothing like how Google works.

Most med spa owners who know about AI search assume it works like Google: rank well, appear more. That's not what's happening. The inputs are different. The sources are different. The ranking logic is different. And the practices that understand this are pulling ahead in a channel that's growing every month.

Here's exactly what happens when a patient asks an AI to recommend a med spa.

Step one: the AI interprets the intent

When a patient types "best med spa for lip filler in Dallas," the AI — whether it's ChatGPT, Perplexity, or Gemini — first decodes what she actually wants. This is a local, transactional query with aesthetic intent. She wants a specific treatment, in a specific city, from a provider she can trust.

That interpretation shapes everything that follows. The AI is not going to return a general article about lip filler. It's going to try to return specific, local, bookable businesses. To do that, it needs to pull from sources it trusts for local business information.

Step two: ChatGPT queries the Bing index

This is the part most people don't know. ChatGPT does not use Google's index. When it searches the web to answer a local business query, it queries Microsoft's Bing. This matters enormously for practices that have invested entirely in Google SEO and never thought about Bing.

A practice with a strong Google presence but no Bing presence is, from ChatGPT's perspective, a practice that barely exists online. The Bing Webmaster Tools dashboard is free. Submitting your sitemap takes ten minutes. Most practices have never done it.

Beyond the Bing index, ChatGPT also draws from:

  • Review aggregators — Google Reviews, Yelp, Healthgrades. Volume, recency, and average rating all factor in.
  • Mentions on trusted sites — local news articles, beauty blogs, medical directories, RealSelf. When multiple credible sources mention the same practice, the AI gains confidence.
  • The practice's own website — specifically, how clearly it describes treatments, credentials, and what makes it distinctive. Thin websites generate thin AI descriptions, or no mention at all.

A typical ChatGPT recommendation for a local med spa sounds something like: "Lumina Aesthetics — a medically supervised practice in Midtown known for natural-looking results, with over 300 five-star reviews and strong mentions across local beauty publications." Every phrase in that sentence came from somewhere. The AI synthesized it from sources it found credible.

Step three: Perplexity searches differently

Perplexity operates more like a real-time search engine than ChatGPT does. It actively crawls the web at query time and cites its sources in the response. For local business queries, it draws heavily from:

  • Recent web content — fresh blog posts, recent Instagram embeds, new press mentions
  • Review platform listings
  • Directory sites that aggregate local business information

The recency weight in Perplexity is significant. A practice that posted on Instagram last week, got a review yesterday, and has a press mention from last month looks very different to Perplexity than a practice with the same underlying quality but a stale online footprint. More than half the med spas we scan haven't posted on Instagram in over a month — which directly hurts their Perplexity visibility.

Perplexity also shows its sources, which means patients see exactly where the information came from. A recommendation backed by three cited sources — a review platform, a local publication, and the practice's own treatment page — carries more weight than one with a single source.

Step four: Gemini leans on Google's ecosystem

Gemini is Google's AI assistant, and it behaves differently from the other two. It draws more heavily from Google's own ecosystem: Google Maps data, Google Reviews, and Google Business Profile information. A practice with a strong, complete GBP tends to surface more reliably in Gemini than one with a sparse profile.

This means the three major AI platforms have meaningfully different inputs:

  • ChatGPT: Bing index, reviews, third-party mentions, website content
  • Perplexity: Real-time web, recent content, review platforms, citations
  • Gemini: Google ecosystem — Maps, GBP, Reviews, Search

A practice optimized for only one of these inputs will appear in some AI responses and not others. A practice with strong signals across all these sources can appear consistently across all three platforms.

Why some practices appear and others don't

The practices that show up in AI responses have, in most cases, not specifically optimized for AI search. They've done the right things across multiple channels — built review volume, maintained a content-rich website, gotten mentioned in local publications, kept their GBP current — and those signals happen to be exactly what AI systems look for.

The practices that don't appear are often strong in one area and weak in others. A practice with excellent Google SEO but no Bing presence, no third-party mentions, and a stale review profile will be largely invisible to ChatGPT. A practice with a great GBP but no recent content and thin reviews will struggle in Perplexity.

Most med spas don't appear in AI responses for their target treatments. In most local markets, that means only a small handful of practices are receiving warm, AI-qualified leads from patients who have already been told "this is a good option." Everyone else is invisible in that channel.

What you can do today

Start with what you can control immediately:

  • Submit your sitemap to Bing Webmaster Tools. Free, ten minutes, immediate impact on ChatGPT visibility.
  • Check that your website has dedicated, content-rich pages for each treatment you offer. AI systems summarize what they find — thin pages generate thin or no mentions.
  • Get your review velocity moving. A consistent flow of new reviews is one of the strongest signals across all three AI platforms.
  • Post on Instagram regularly. Perplexity weighs recency heavily, and a dormant account is a signal you're not currently active.
  • Claim and complete profiles on Healthgrades, RealSelf, and Yelp. These are the third-party sources AI treats as credible references.

Before doing any of that, it's worth knowing your current baseline — which channels you're visible in, which you're not, and what the specific gaps are. Most practices are surprised by what they find.

Run your free Cornflower Scan at cornflower.ai/scan — it shows your AI visibility score alongside your full five-channel breakdown, so you know exactly what the AI sees when a patient asks about med spas in your city.