Generative engine optimization · Original data
We ran the same AI search across six local trades in ten cities. Only three answers named a real business.
We asked ChatGPT the kind of question a real customer types, in ten different cities, across med spas, dentists, salons, gyms, and personal injury lawyers, sixteen runs in total. Thirteen came back built entirely from directories, review sites, and listicles. Not one business's own website made the cut.
We work with med spas and salons, so that's where our data runs deepest. But we cast the net wide on purpose, because the real question isn't whether your spa is invisible. It's whether that's a spa problem at all, or something larger that spas simply sit inside.
What we ran
GPT-4o, with live web search enabled, one question per city: "best med spa in Miami," "best hair salon in Tulsa," and so on. Sixteen runs across August 2026, covering med spas, dentists, hair salons, gyms, personal injury lawyers, plus one dermatology check to see how a more trust-sensitive medical field fares. For each answer we logged every cited source and counted how many came from the business itself rather than a third party talking about it. This is one run per query, not an average across repeated runs, so treat the numbers as directional rather than lab-grade.
Med spas and salons, up close
Best med spa in Miami pulled five sources and named zero clinics. We reran the question in Austin, Phoenix, Chattanooga, and Boise. Four of the five med spa runs came back the same way. Boise broke the pattern, where spa35.com appeared alongside three directories.
One thing held across every med spa city we checked: medspascout.com showed up in all five runs, and medicalspalocator.com showed up in four of the five. Whatever lands a med spa on those two listings seems to outweigh anything on its own site.
Hair salons mirrored this almost exactly. Miami's best-hair-salon answer drew eight sources and named zero salons. Tulsa matched it. Chicago was the outlier, where bangsalonchicago.com made the list alongside three general local guides.
| City | Trade | Sources | Business sites named |
|---|---|---|---|
| Miami | Med spa | 5 | 0 |
| Austin | Med spa | 7 | 0 |
| Phoenix | Med spa | 4 | 0 |
| Chattanooga | Med spa | 9 | 0 |
| Boise | Med spa | 4 | 1 (spa35.com) |
| Miami | Hair salon | 8 | 0 |
| Tulsa | Hair salon | 5 | 0 |
| Chicago | Hair salon | 4 | 1 (bangsalonchicago.com) |
This wasn't specific to us
We added dentists, gyms, and personal injury lawyers to test whether this was specific to spas and salons or something structural. It wasn't specific to us. Dentist answers in Denver and Nashville were built from Zocdoc, RateMDs, WebMD, and the American Dental Association's own directory, with zero practice websites in either city. Personal injury lawyer answers in Miami and Denver came entirely from Avvo, Justia, SuperLawyers, and a Forbes roundup, again with zero firm sites. A gym search in Seattle pulled from Time Out, two gym directories, and Reddit threads, with no gym's own site listed.
The one run that broke the pattern was dermatology, run once in Miami as a check on a more trust-heavy medical field. Two of the four sources were the clinic's own site and a university medical center's directory page. That's the closest thing to a genuine exception in the whole set, and it points at something specific: institutional weight, not simply being a medical practice, is what got through.
The three that made it
Across sixteen runs, three answers included a business's own website: the Boise med spa, the Chicago salon, and the Miami dermatology practice tied to a university medical center. The three shared nothing obvious in size or specialty. What they had was a form of established weight the other thirteen lacked, whether that came from years of local press, an institutional affiliation, or simply a smaller market with fewer directories crowding the same slot.
City size alone didn't decide the outcome. Boise, a small market, produced one of the three exceptions. Chattanooga, a similarly sized city, produced nine sources and zero business websites.
Reddit kept surfacing
Reddit appeared in the source list for five of the ten additional city runs: Austin, Denver, Nashville, Seattle, and Chattanooga. It wasn't a minor mention either, it sat among the directories the engine leaned on most. If your business gets talked about in a local subreddit thread, that thread has a real chance of feeding an AI answer about you, whether or not you've ever posted there yourself.
What we'd check first
If you run a med spa or salon, start with the two or three directories that keep recurring across cities in your category, not just Google Business and Yelp. For med spas that's specifically medspascout.com and medicalspalocator.com. Make sure your name, address, and phone number match exactly across every listing, since a mismatch reads as two smaller, less-established businesses to a model rather than one consistent one.
Then look at what's already being said about you off-site: local press, Reddit threads, and review platforms with consistent details. That's the layer these thirteen answers were built from, and it's the layer most owners never think to check.
The thirty-second version of that check is free. Search your business name and city and see exactly what ChatGPT and Perplexity say about you right now.