When someone asks an AI assistant which dentist to see, specific practices get named. We measure which ones.
Of the 29 Sydney implant and cosmetic practices we tracked, 14 were never named in any of the 116 market answers from ChatGPT, Gemini, Claude and Perplexity. Six of those 14 had their own website cited in answers that named other practices.
Fig. 1 — Sydney panel, 26–27 August 2026. Your report shows your own city, your own numbers.
The audit
- Your presence count across the 116 market answers, with the stored, timestamped answers behind every count
- The verified clinics named most often in your city, with their counts
- The sources each assistant cited in market answers — and, separately, the sources it cited when asked about you by name
- What the assistants say when asked about your practice by name — 8 answers, checked for accuracy against your own website
- A prioritised 90-day roadmap your web team can execute from
Received an email from me? Reply to it and I’ll send your one-page snapshot.
The audit stands on its own. If you later want the panel re-run to measure change, or help executing the roadmap, that’s available afterwards — nothing to cancel, no lock-in.
How to read the numbers
The panel uses 40 patient-style question templates. For market visibility we measure 29 Sydney market questions across four assistants — 116 market answers — and count the answers in which each clinic is named. Separately, each assistant is asked about your practice by name — 8 answers — to see what it says once it knows who you are.
Results are counts, not rankings. Until repeat runs establish normal run-to-run variation, small differences between practices should not be read as meaningful.
The four assistants don’t cite the same sources
In the 116 Sydney market answers, one directory was cited in 22 of ChatGPT’s 29 answers and in none from Claude, Gemini or Perplexity. The source cited in the most answers — 46 of the 116 — was a set of “best providers in Sydney” pages published by a clinic in Townsville. Checking one assistant tells you about one assistant.
| Source | Answers citing it (of 116) | ChatGPT (of 29) | Claude (of 29) | Gemini (of 29) | Perplexity (of 29) |
|---|---|---|---|---|---|
| townsvilledental.clinic | 46 | 12 | 12 | 1 | 21 |
| whatclinic.com | 31 | 7 | 10 | 5 | 9 |
| bestinhood.com | 22 | 22 | 0 | 0 | 0 |
Fig. 2 — Sources cited in the 116 market answers of the Sydney panel, 26–27 August 2026 (29 questions × 4 assistants). Counts are answers in which the source was cited at least once. Three of 332 domains cited. Individual practice websites are excluded from this sample.
Why not just check ChatGPT yourself?
You can, and you should — it takes a minute. What one check can’t tell you is whether that answer is typical, how the other three assistants differ, or which sources it cited. A single check on a single assistant on a single day is a data point. Twenty-nine market questions across four assistants — 116 answers — is a measurement.
Who runs it
I’m Eugene Naude. FIDESTIS is me — one person who builds and runs the measurement, writes the report, and takes the call.
I built this because what AI assistants say about a practice is measurable and almost nobody is measuring it at practice level. Analytics can’t see it: there’s no click to record when an assistant names a clinic. So I built the instrument that can.
I work with a small number of practices at a time, in Australia and the UK.
eugene@fidestis.com
What I don’t claim
No one can guarantee that an AI assistant will recommend a particular practice, and anyone who tells you otherwise is selling something they can’t deliver. What I sell is measurement: which practices the assistants named, how often, which sources they cited, and a roadmap.
The same panel can be re-run so change is measured rather than asserted — and called real only once repeat runs have established normal run-to-run variation. AI answers vary between sessions, which is why every figure in your report is tied to a timestamped response you can inspect.
Next step
Received an email from me? Reply to it and I’ll send your one-page snapshot for your city.
Official FIDESTIS outreach may come from visibilityfindings.com or presencefindings.com. fidestis.com is our main website and business domain. If an email claiming to be from us comes from anywhere else, it isn’t.