November 18, 2026

AI Visibility Study 2026: Which Metro Manila Clinics Do ChatGPT, Gemini and Perplexity Recommend?

November 18, 2026

Last updated: 18 November 2026 · Data collected 13 and 18 November 2026 · By Dominic Barrios, Bots at Work

Short answer: We asked ChatGPT, Gemini and Perplexity ten patient questions about dental and aesthetic clinics in Metro Manila, twice, and read all 60 answers. No clinic was named in more than 32% of them (the most-named clinic appeared in 19 of 60), five of the 20 clinics we followed were never named by ChatGPT, and almost all of ChatGPT’s recorded sources were Google Maps listings (131 of 133). Among the 13 clinics whose Google Maps listing we could confirm, the six that were among the seven most-named had a median of 240.5 Google reviews, against 164 for the other seven. That sample is small and review counts differ by branch, so treat it as a hint, not a finding.

If you run a clinic, the practical point is simple: which AI tool a patient uses changes which clinics that patient sees. ChatGPT gave about 7 names per answer, Perplexity about 9 and Gemini about 17. This study shows what we measured, which clinics were named, and what the most-named clinics have in common.

Key findings

  1. The lists are fragmented. Across 60 answers the 20 clinics were named 170 times in total, an average of 8.5 answers each. Even the most-named clinic was missing from two answers in three.
  2. ChatGPT has the shortest list. Of those 170 mentions, 37 came from ChatGPT, 58 from Perplexity and 75 from Gemini. Thirteen of the 20 clinics were named by all three tools; five were never named by ChatGPT.
  3. ChatGPT’s sources were Google Maps listings. In the clinic answers, 131 of the 133 sources we recorded for ChatGPT were Google Maps listings. In our separate 13 November 2026 check of AI Visibility questions, 12 of 12 were too. For these local questions, a Google listing looks like the main source ChatGPT used.
  4. Perplexity leaned on one directory. One clinic-directory website supplied 93 of the 200 sources we recorded for Perplexity, nearly half. For Gemini, every one of the 294 sources we recorded (up to 15 per answer) was a Google redirect link, so we cannot see which pages it used.
  5. Reviews are a hint at best. Among the 13 clinics whose Google Maps listing we could confirm (see the method), the six that were among the seven most-named had a median of 240.5 reviews, against 164 for the other seven. But the pattern did not hold for ChatGPT: the three confirmed clinics it never named had a median of 287 reviews, against 185.5 for the ten confirmed clinics it did name. And one clinic with 563 reviews was named in only 3 of 60 answers. With samples this small, reviews alone do not explain who gets named.
  6. Structured data was rare, and slightly more common among the leaders. Three of the seven most-named clinics had clinic or FAQ structured data on their homepage (anything beyond a plain “Organization” label), against 2 of the other 12 websites we could check. Only 2 of the 19 websites we could check blocked an AI crawler, and one of them was still named by ChatGPT five times, so blocking did not keep that clinic out of ChatGPT’s answers.

The 10 most-named clinics

Ordered by the number of answers that named the clinic (ties ordered by score, then alphabetically). “Named” means the clinic appeared in the list of suggestions in an AI answer; it is not a rating of quality. The other 10 clinics were named between 3 and 7 times; we show only the count.

#ClinicAnswers naming it (of 60)ChatGPTPerplexityGeminiScore (0-100)
1Affinity Dental Clinics1956852
2Belo Medical Group1846853
3GAOC (Gan Advanced Osseointegration Center)1525846
4Aivee Clinic1545643
5CorDerm Advanced Dermatology and Laser Center1207533
6Urban Smiles Dental Clinic1253433
7Dental World Manila1042428
8Cosmo Dental921623
9Gerochi Dental & Implant Center814322
10Elevate Dental714220

Tenth place was a tie at 7 answers; Elevate Dental is shown because it has the higher score. The score is our own: the share of the 60 answers that named the clinic, nudged up for being listed early.

A few things worth noticing. The clinic Perplexity named more than any other (CorDerm, 7 times) was never named by ChatGPT, which is why checking one tool is not enough. Belo has the highest score (53) even though it was named once less often than Affinity, because it was listed earlier in the answers (average position 3.6, against 5.7). And the middle of the table is close: the difference between 4th and 7th place is five answers out of 60.

What the most-named clinics do differently

These are patterns in our sample of 20 clinics and 2 rounds, not guarantees. Treat them as a checklist to test, not a recipe.

  1. A complete Google listing with genuine reviews. Google Maps listings were 131 of the 133 sources we recorded for ChatGPT. Fill every field, keep branch details consistent, reply to every review, and ask real patients the same way after each visit. Never pay for reviews or offer rewards for them; Google forbids it.
  2. A homepage that says what the clinic is. Add clinic or business structured data and FAQ markup so machines can read who you are, what you treat and where.
  3. A page that answers patient questions directly. The AI answers cited some clinics’ own websites as a source (Belo’s 10 times, Affinity’s 7 times, out of 60 answers). Pages that answer “how much”, “how long” and “who is it for” in the first two sentences give an AI something to quote.
  4. Open doors for AI crawlers, unless you mean to close them. Check your robots.txt. In our sample, blocking did not stop ChatGPT from naming one clinic, but it can only hurt where a tool reads your site.
  5. Check all the tools, not one. Ask ChatGPT, Gemini and Perplexity your patients’ questions every month, and write down who is named. Our guide to how to check your AI visibility shows how.

To see how each tool picks businesses, with our own dated data, read how ChatGPT, Gemini, Perplexity and Copilot choose which businesses to recommend.

How we ran the study (method)

  • Dates: two rounds, 13 November 2026 and 18 November 2026.
  • Tools: ChatGPT, Perplexity and Gemini, each with web search switched on, reached through an API rather than the consumer apps. What an individual patient sees on their phone can differ.
  • Questions: ten, asked word for word to every tool in both rounds (3 tools x 10 questions x 2 rounds = 60 answers):
  1. What is the best dental clinic in Makati?
  2. Where can I get teeth whitening in BGC Taguig?
  3. Which dental clinic in Quezon City is good for affordable braces?
  4. What are the most trusted dental clinic chains in Metro Manila?
  5. Where should I get dental implants in Manila?
  6. What is the best dermatology clinic in Manila for acne?
  7. Which aesthetic clinic in Metro Manila is best for laser hair removal?
  8. What are the most trusted aesthetic and skin clinics in the Philippines?
  9. Where can I get Botox or fillers safely in Makati or BGC?
  10. Which skin clinic in Quezon City is affordable and good?
  • Choosing the 20 clinics: we took the 20 clinic brands named most often in the first round (30 answers; several clinics tied at two mentions, so we included 20 in total), merged duplicate names, and removed non-clinics such as booking sites. Both rounds were then counted for those 20.
  • Counting: for each clinic we counted the answers that named it (out of 60), recorded the tool, its position in the list, and whether its own website was cited. Names were read from the answer text; we matched them to the 20 clinics. Sources are the links each answer gave, recorded up to 15 per answer.
  • Site checks (18 November 2026): for each clinic we looked up the first Google Maps result for its brand (rating and review count), read the website’s robots.txt for AI crawler blocks (GPTBot, Google-Extended, ClaudeBot and others), and read the homepage for structured data. One clinic’s Maps result had no website, so those checks cover 19 of the 20.
  • Checking the Maps matches: most clinics have several branches with very different review counts, and Google does not always list the same branch first. So we ran a second, more specific Maps lookup (“clinic name” plus “Philippines”) on 18 November. A clinic’s review count is used only if the second lookup returned the same location with the same review count (within 1). That was true for 13 of the 20 clinics. For the other seven, the first listing changed between lookups, so we do not use their review counts anywhere in this study. Review figures are for one location of each clinic, not the whole brand.
  • Score: 0 to 100, based on the share of answers that named the clinic, nudged up for being listed earlier.

Limits you should know about

  • Small sample. Twenty clinics, two rounds five days apart. AI answers change from day to day.
  • We started from clinics AI already names. The 20 clinics came from the first round’s answers, so this is a study of clinics that AI mentions, not of every clinic in Metro Manila.
  • Reviews are one location, and for 13 clinics only. The review comparison uses the 13 clinics whose Google Maps location we could confirm, and one location per clinic. Six of the seven most-named clinics are in that group, so the comparison covers 6 against 7.
  • Correlation, not cause. Reviews, structured data and crawler access sometimes went with being named, but we have not shown that they cause it.
  • “Named” is not “best”. This study measures how often AI tools name a clinic. It says nothing about clinical quality.
  • The tools are a moving target. We will say so on this page if we repeat or update the study.

Ask for your clinic’s scorecard

If you would like to see how ChatGPT, Gemini and Perplexity describe your clinic, and what to fix first, ask for your clinic’s scorecard: Book an AI Visibility call. You can also read about our AI Visibility service.

Frequently asked questions

Which clinic do ChatGPT, Gemini and Perplexity name most often?

In our 60 answers, Affinity Dental Clinics was named most often (19 of 60), followed by Belo Medical Group (18 of 60). Even so, neither appeared in more than a third of the answers, and the order changed by tool.

We took the 20 clinic brands named most often by the three AI tools in the first round on 13 November 2026, merged duplicate names, removed non-clinics such as booking sites, and counted both rounds for those 20.

Not clearly. Among the 13 clinics whose Google listing we could confirm, six of the seven most-named had a median of 240.5 reviews against 164 for the other seven, but the three confirmed clinics ChatGPT never named had a median of 287. One clinic with 563 reviews was named in only 3 of 60 answers. Reviews should come from real patients, never from payment or rewards.

In our answers ChatGPT gave about 7 names per answer, Perplexity about 9 and Gemini about 17. Gemini gives long lists; ChatGPT gives a short one, so getting onto ChatGPT’s list is harder.

No. The study counts how often AI tools name a clinic. It does not measure treatment quality, safety or price.

Ask for your clinic’s scorecard. It shows how the three AI tools answer your patients’ questions and what to fix first. Use the booking link in this article.

About Bots at Work

Bots at Work is an AI Visibility agency that helps businesses appear in AI-generated answers on ChatGPT, Perplexity, and Gemini — not just on traditional Google. Founded by Dominic Barrios, the company applies AEO, GEO, and LLM SEO strategies proven on its own business before recommending them to clients. Bots at Work operates out of Ortigas Center, Pasig, and serves businesses in the Philippines and internationally through botsatwork.ph. To learn how AI Visibility can get your business cited by AI models, visit botsatwork.ph/contact-us.