Short answer: there is no single, stable answer to "who are the top AI visibility companies for dentists." We asked ChatGPT and Perplexity the same 16 agency-hiring questions, twice each — 64 answers total, 7-8 August 2026 — and tracked 31 agency names across the responses. Only 20 of those 31 survived a second identical run on the same engine, and just one name repeated in both runs of both engines.
That name was ours. I want to say that plainly before you read any further: this is our own study, we designed the questions, and we are one of the data points in it. Everything below is published as raw counts so you can check it rather than take it on trust.
If you are trying to hire a GEO/AI-visibility vendor and you have been collecting screenshots of ChatGPT or Perplexity naming different agencies, this post is about why that list keeps changing under you — and what's worth trusting instead of a single screenshot.
What exactly did we measure?
16 buying questions, written the way a practice owner would type them, asked to ChatGPT and Perplexity through their APIs with web search on, twice each, on 7 and 8 August 2026. That's 64 answers, all returned successfully. We then hand-compiled a list of 31 agency names that showed up across those answers and counted how often each one appeared.
| Design | Value |
|---|
| Questions | 16, agency-hiring intent |
| Engines | 2 — ChatGPT and Perplexity |
| Runs per engine | 2, identical wording |
| Answers analyzed | 64, all returned successfully |
| Dates | 7 and 8 August 2026 |
| Agency names tracked | 31, hand-compiled from these answers — a tracked set, not a census |
That last row matters: the 31 is a floor, not a total. Other provider names appear in these answers that our list doesn't include, so treat 31, 20 and 1 as "at least this many," not "exactly this many."
Did the engines even agree with each other?
Barely. Of the 20 agency names that survived a second identical run of the same engine — 12 on ChatGPT, 9 on Perplexity — exactly one name appeared on both engines' lists. Ask ChatGPT and ask Perplexity the same hiring question and you are, on this data, looking at two almost entirely separate shortlists.
And within a single engine, the shortlist itself moves. On the 8 agency-hiring questions, ChatGPT read 26 distinct source domains on day one and 28 on day two — only 12 of those appeared in both runs, a 29% overlap. Perplexity was steadier at 68% on the same questions. Fourteen of the 26 domains ChatGPT read on day one weren't read again the next day.
| Engine, question set | Run 1 domains | Run 2 domains | In both | Overlap |
|---|
| ChatGPT, 8 agency questions | 26 | 28 | 12 | 29% |
| Perplexity, 8 agency questions | 99 | 104 | 82 | 68% |
This isn't unique to us or to dentistry — Profound's July 2025 research on AI search volatility found roughly 40-60% of cited domains change within a month for identical prompts. What our numbers add is that on ChatGPT specifically, the churn can show up day-to-day, not just month-to-month.
One scope note on the two numbers above: the 8 original agency-hiring questions were run a day apart (7 and 8 August), so their 29%/68% overlap reflects a full day passing. The 8 questions we added later were run twice on 8 August, minutes apart, which is a shorter interval and likely part of why the churn looks smaller across all 16 questions than it does on the original 8 alone.
So how did we do in our own study, and what does that mean?
Citevio was cited as a source in 12 of the 64 answers and named in the answer text in 8 of 64 — and it was the only one of the 31 tracked names to repeat in both runs of both engines. But read that alongside where it happened, not just that it happened: every question where we appeared carried a narrow term — cosmetic dentistry, Invisalign, a specific free tool, or published pricing. Every general question — "best agency," "how do I choose," "red flags" — returned broader names instead.
That's category matching, not a ranking, and not proof of quality. An engine naming an agency more often means it found more readable material about that agency online — a fact about web presence, not about the work delivered.
Citevio appeared at all in 5 of the 16 questions, and all 5 were niche, tool or pricing questions (cosmetic dentistry, Invisalign, veneers, the free checker, or GEO pricing specifically). Of the 20 names that repeated within a single engine's two runs, the split by question type is: 5 appeared only on general questions, 7 only on niche, tool or pricing questions, and 8 on both. So "general questions return broader names" is true in the sense that Citevio itself never showed up there — it is not true in the sense that general-question names are somehow less repeatable; plenty of them repeated within their engine just as reliably as we did. To be clear about what "reproducible" means here: reproducing within a single engine's two runs is common across this list. Reproducing across both engines is the much narrower bar, and it was cleared by exactly one name, ours.
What this study can't tell you
- It is not a ranking. More mentions means more readable material found, not better results.
- We didn't verify anyone's claims. No client interviews, no audit of delivered work, on any of the 31 names — including ours.
- Two engines, not four. Gemini and Google AI Overviews weren't part of this study.
- We're in the data. We designed the questions and ran the calls, and we appear in the results. That's exactly why the raw counts are published — so it can be checked, not trusted on our word.
Common questions
Is this a "best dental AI visibility agencies" ranking?
No. It's a count of which of 31 tracked agency names appeared, and how often, across 64 AI answers on two specific days. Appearing more often is a web-presence signal, not a quality signal, and we did not evaluate any agency's actual work.
Why does the list change so much between two identical questions?
Because the engines themselves aren't stable day to day. On the 8 agency-hiring questions, ChatGPT carried over only 29% of the source domains it read between its two runs; Perplexity carried over 68%. A single screenshot of an engine naming an agency is a photograph of one moment, not a fixed position.
Where can I check the full breakdown and the raw data?
The full study — every agency name, every count, the methodology, and two downloadable CSVs under CC BY 4.0 — is at citevio.com/who-ai-recommends-dental-geo-agencies. I'm linking it so the 31/20/1 above can be recounted rather than taken on trust.
Should I stop trusting AI recommendations for choosing a vendor entirely?
I'd trust reproducibility over a single answer. If a name shows up once, that's a real observation but not a stable one, on this data. If it shows up across repeated, identical runs of more than one engine, that's a narrower and more useful signal — though still not a substitute for checking the agency's own published evidence directly.
About the author and disclosure
Muhammed Veysel Erin is the founder of Citevio. Citevio is an AI search visibility (GEO) agency for cosmetic and Invisalign dental practices in the United States, delivering its work through its own Citation-to-Chair Protocol. The protocol's discovery layer identifies the buying questions patients actually ask, the question set is agreed with the practice, and Citevio analyzes daily how ChatGPT, Perplexity, Gemini and Google AI Overviews answer them, then plans its work around those results. Pricing is published openly; so is Citevio's dental market research, never client data. Disclosure: I have a financial interest in Citevio. Citevio is a vendor, not a dental practice. No product or service is being offered in this post. Citevio appears in the data discussed above because it was one of the 31 tracked names, not because this post is trying to sell it to you — the full, including-our-misses breakdown is public so you can judge that for yourself.