71% of Body Contouring Spas Are Invisible to AI Search. Here Is the Data.

The short version: On October 22, 2026, we asked ChatGPT and Gemini to recommend a body contouring spa in each of the 20 largest US cities, three times each. That is 120 answers. Of the 312 spas and med spas Google Maps lists first for "body contouring spa" in those cities, 223 (71%) were never named. For day spas and other non-medical spas it was 85 of 109 (78%). Spas the apps named in most answers had a median of 179 Google reviews. Spas never named had 57.

Do AI apps recommend body contouring spas?

Yes, but almost always the same few. When we asked ChatGPT and Gemini to recommend a body contouring spa in the 20 largest US cities, 223 of the 312 spas and med spas on the first page of Google Maps (71%) did not come up in any of the 6 answers for their city. Only 3 spas (1%) were named every time.

The problem is not that AI will not recommend a body contouring spa. It is that it keeps sending people to a small group of spas, and most spas are not in it.

Any single round of asking left out 81% to 85% of spas. Asking three times turned up a few more names, which is why the combined number is 71%. We report the combined number because it is the harder one to be left out of.

How did we test this?

We took the first 20 businesses Google Maps lists for "body contouring spa" in each of the 20 largest US cities, then asked ChatGPT and Gemini to recommend a body contouring spa in each city. We asked each app three times per city on October 22, 2026, logged out, in a fresh session every time. That gave us 120 answers, and we checked every one for which of those businesses it named.

The question, word for word: "recommend a body contouring spa in [City], [State]"

The apps: ChatGPT and Gemini, both used logged out. Perplexity, which we included in an earlier cryotherapy study, now refuses to answer unless you sign up, and Google's AI Mode blocked our logged-out requests as unusual traffic. If and when we can measure Google's AI answers the same way, this page will be updated in place and the engine list above will change.

The cities: the 20 largest US cities by population (2020 Census): New York, Los Angeles, Chicago, Houston, Phoenix, Philadelphia, San Antonio, San Diego, Dallas, San Jose, Austin, Jacksonville, Fort Worth, Columbus, Indianapolis, Charlotte, San Francisco, Seattle, Denver and Oklahoma City. We used each city's own name. Asking about a nearby suburb can return a different list.

Which businesses we counted: the first 20 non-sponsored businesses Google Maps lists for "body contouring spa in [City], [State]", in Maps' own order, pulled the same day. Nineteen cities gave 20; Indianapolis listed only 11. That is 391 businesses in all. Maps mixes in weight loss clinics, massage therapists and skin clinics for this search, so we recorded each business's Maps category and report the 312 whose category is a spa category (Medical spa, Spa, Day spa, Health spa, Massage spa, Facial spa, Spa and health club) as the headline. Across all 391, the share never named was 72%, so the cut does not change the picture.

What counted as named: the business's name appeared in the answer, in the text or in a place card the app showed inside the answer. Matching was scripted on normalized names and then checked by hand three times, including a sweep of every never-named business for near misses. A chain mentioned by brand only was credited to every location of that brand in the city's sample, which makes "never named" smaller, not larger.

Why logged out, in a fresh session: so past chats and personal settings could not change the answers. That is close to what a new customer sees, though someone signed in may get a different list.

Why three times: these apps give different answers to the same question from one try to the next. One check proves very little, so a spa only counted as never named if it came up in none of its 6 answers.

Does ranking high on Google Maps get a spa named by AI?

Not reliably. Every business in this study is one Google itself already puts on the first page for body contouring in its city. Of the 46 spas in the top three Maps positions, 21 (46%) were never recommended by either app. Of the 17 spas Maps listed first, 8 were never named.

A high Maps position helps, since the never-named share is lower at the top than across the whole sample, but it is nowhere near a guarantee. Google Maps and the AI apps are different systems. One ranks mostly on distance, category and reviews. The others pull from whatever they read about your business, and the two lists overlap less than most owners assume.

The second column comes from a rougher count than the rest of this study, so read it as a direction rather than an exact figure. Even so, it points one way: a large share of what both apps recommended were businesses outside Maps' first 20 for this search. The apps are not simply reading the Maps list back.

Which body contouring spas do get recommended?

The established ones. Spas named at least once had a median of 100 Google reviews. Spas named in 4 or more of their 6 answers had 179. Spas never named had 57.

Reviews alone are not enough. 44 of the 223 spas that were never named had 179 or more Google reviews, as many as the typical spa the apps named most often. Of the 118 spas with 100 or more reviews, 73 (62%) were still never named. Being well reviewed does not put a spa in the answer by itself.

Day spas are left out more often than med spas. Med spas were never named 68% of the time. Day spas and other non-medical spas, 78%.

Which cities are hardest?

None of them were easy. The share of spas never named ran from under half in Indianapolis, which had the shortest Maps list, to over 90% in San Diego. The apps name about the same small number of spas per answer no matter how many exist, so in a crowded city each spa's chance of a slot is smaller.

These are counts of the spa-category businesses on Maps' first 20 for each city, so the sample per city is small and a single spa moves the percentage. Read the column as a pattern, not a league table.

Why does the same question give different answers?

Because these apps do not keep a fixed list. We asked each app the same question three times on the same day, logged out, in a fresh session. ChatGPT named the same set of spas from the Maps sample in all three rounds in only 2 of 20 cities. Gemini did in 3 of 20. And 18 of the 120 answers named nobody from Maps' first 20 at all. On average an answer named 1.9 businesses from the sample.

What stayed steady was the overall picture. In each single round, 81% to 85% of spas went unnamed. The pattern is stable. Which spas fill the slots is not.

That has a practical consequence for any owner checking on themselves. Being named once is not a position, and disappearing once is not a penalty. A single check proves almost nothing either way. It is why we asked 6 times per spa, and why a spa only counted as never named if it failed to come up in all 6.

Both apps did about equally badly by the spas. Each one, on its own, never named roughly four spas in five across its three answers.

What can a spa owner do about it?

These are the things the named spas had in common. They are patterns, not promises. Nothing here guarantees an AI app will name any particular business, and we did not record where each app got its answers in this run, so we cannot tell you which source to fix first.

1. Say the words a buyer would type, on the website and on the Google listing. "Body contouring", the treatment names you actually offer, and your city, in the page title, the main heading and the Google Business Profile services list. A spa that offers body contouring but never writes the phrase down is invisible to a system that works by reading.

2. Fill in the Google Business Profile completely. Every service as its own entry, a real description, correct hours, a booking link, current photos. Gemini shows Google place cards inside its answers, so the listing is a large part of what it has to go on.

3. Ask every customer for a Google review. Spas named in most answers had a median of 179 reviews. Spas never named had 57. Ask everyone, not only the happy ones, and never offer anything in return. Reviews are not a switch that turns AI recommendations on, but the spas the apps name consistently are the ones with a lot of them.

4. Check it yourself, more than once. Ask each app to recommend a body contouring spa in your city, three times, in a fresh session. One answer tells you nothing, as the section above shows. Write down who was named instead of you, and look at what their website and listing say that yours does not.

5. Answer the leads you do get, fast. This study is about being found. Being found is wasted if the inquiry sits unanswered. The spas that win on both are the ones that treat the first reply as part of the service.

One honest note on what is realistic. In a city where Maps lists 20 spas for the search and the apps name two per answer, most spas will never be named at all, and no amount of website work changes how many slots exist. The winnable goal is the specific question, the named treatment, the neighborhood, the "near me" with a reason, rather than the broad one. Those are reachable in months. The broad one is reachable for a handful.

What this study cannot tell you

Everything above comes from one day of asking. Here is what that does and does not support.

It is a snapshot, not a trend. All three rounds ran on October 22, 2026, within about an hour. That shows how much answers move within a day. It says nothing about how they move week to week. We will repeat it, and this page will be updated in place.

We could only measure two apps. Perplexity, which we included in an earlier cryotherapy study, now refuses to answer unless you sign up, and Google's AI Mode blocked our logged-out requests as unusual traffic. If and when we can measure Google's AI answers the same way, this page will be updated in place and the engine list above will change.

We asked while logged out. That is close to what a new customer sees, and it keeps personal chat history out of the results. Someone signed in may get different answers, and logged-out users get a lighter version of some apps.

312 is a sample, not a census. It is the spa-category businesses among the first 20 Google Maps lists for "body contouring spa" in each city. Google Maps returns well over 100 businesses for that search in a large city, so this is the top of Google's own list, not every spa in town. Maps order in a signed-in browser can differ slightly from what a stranger sees; we pulled the list once, read-only.

We asked one question, in one wording. Different words produce different lists. "Body sculpting", "CoolSculpting near me" or "non-surgical fat reduction" will each return something different.

We used city names. A nearby suburb can return a different list, and a spa that never came up for its big city may do well for the town it actually sits in.

Matching was scripted, then hand-checked. One person reviewed the near misses and the never-named list. There was no second reader.

These are patterns, not causes. Spas the apps name have more reviews and sit higher on Maps. That does not prove either one causes the naming, and 44 spas with 179 or more reviews were never named at all.

This is not the same study as our cryotherapy one. We ran a similar check on cryotherapy centers in September 2026 and landed on a similar share. The two used different apps, different search frames and different samples, so treat them as two separate snapshots, not one trend.

Frequently asked questions

Does this mean my spa is invisible on Google too?

No. These are different systems. Every spa in this study is already on the first page of Google Maps for body contouring in its city. Our study measured only what ChatGPT and Gemini say when asked for a recommendation, and 71% of those first-page spas were never named.

How often do these answers change?

Constantly. Asked three times in one hour, ChatGPT gave the same set of spas from the Maps sample in only 2 of 20 cities, and Gemini in 3 of 20. Treat any single check as one sample, not a verdict.

Can I pay to appear in these answers?

Not that we could find. None of the 120 answers labeled any recommendation as sponsored. What the apps named was businesses with complete listings, clear websites and plenty of reviews.

Will more Google reviews get my spa named?

More reviews go with being named, but they do not guarantee it. Spas named in most answers had a median of 179 reviews against 57 for spas never named. Still, 44 spas with 179 reviews or more were never named at all. Reviews matter and they are not a switch.

I run a day spa, not a med spa. Does this apply to me?

Yes. Day spas and other non-medical spas were never named 78% of the time, against 68% for med spas.

How do I check my own spa?

Open ChatGPT and Gemini, and in a fresh session ask each one to recommend a body contouring spa in your city. Do it three times per app. Write down whether you were named and who was named instead. Or let us run it for you.