Getting named
The question your customer asks, and why it is the one you can win
Published · Updated Lees dit in het Nederlands
"Best hair transplant clinic Amsterdam" gets four names from ChatGPT, ordered by star rating, with no verdict. "Which Amsterdam clinic is best for FUE with minimal scarring" gets two names and a winner, with a reason. The second question is the one a patient types the night before they decide, and it is the one a practice can be named for without 500 reviews. We ran both on 17 September 2026 and logged what came back.
Three questions, one afternoon
Same assistant, same day, fresh chat each time, logged out, no history. The prompts are the playbook wording we use for every clinic run.
| Question | ChatGPT named | Verdict |
|---|---|---|
| best hair transplant clinic Amsterdam | 4 clinics, ordered by Google rating | none: "shortlist these" |
| I live in Amsterdam Noord and want a hair transplant, where should I go | 5 clinics, each with a district, a Noord clinic placed first | none: "have 2 to 3 consultations" |
| which Amsterdam clinic is best for FUE hair transplant with minimal scarring | 2 clinics | yes: one "stands out" |
On the generic question ChatGPT listed Amsterdam Hair Institute, Medikliniek, Estefarya and Alpha Haarkliniek with their ratings and review counts, then told the user to ask seven questions at each. A patient reading that has a beauty contest and a homework list.
Add one fact, "I live in Amsterdam Noord", and the list changes. A clinic on Kraanspoor that did not appear before moved to the top, "particularly convenient for you", and the other four each got a district. Perplexity, given the same sentence, put a clinic in Landsmeer first instead, on distance. Two assistants, one neighbourhood, two different nearest clinics.
Add a procedure and a worry, "FUE with minimal scarring", and ChatGPT's list halved. Two clinics. One of them "stands out".
Why the specific question has a winner
ChatGPT gave its reason, and the reason is text:
It specifically describes its FUE-CT protocol as emphasizing donor-area preservation, limits the number of procedures per day, and lists several BIG-registered doctors, including Dr. E. Celiker, who is an ISHRS member.
Every clause in that sentence is something the clinic wrote on its own website. Named doctor. Registration. A stated limit on procedures per day. The phrase "donor-area preservation", which is the clinical name for the thing the patient called "minimal scarring". The clinic answered the patient's question before the patient asked it, and the assistant matched the two.
The generic question cannot be won that way, because nothing in "best clinic Amsterdam" gives the model anything to match. It falls back on ratings, and 512 reviews beats 73 whatever your surgeon does. The specific question can be won by whoever has written the specific sentence.
What the assistant is doing
Both assistants, on all three questions, ended the same way: book two or three consultations, and ask these questions when you get there. The model does not choose the clinic. It builds the shortlist the patient takes to the consultation, and the consultations decide.
Being on that shortlist is the whole game. Off it, the patient does not call. On it, your consultation does the rest, which is the part you were already good at.
The questions your customers are asking
A clinic's customers are not typing "best clinic". They are typing what is wrong, where they are, and what they are afraid of. FUE with minimal scarring. Filler that does not look done. A dentist in Zuid who takes nervous patients. An employment lawyer for an expat with a dismissal case. Each of those is a question with a short answer, and the short answer can be you.
The work is to list those situations, write the sentence that answers each one in the words the patient uses and the words the clinician uses, and put that sentence where the assistants read: your own site first, then the registers, directories and review pages they cited on the way to their answer. In the run above those were KliniekErvaringen, Amsterdams Beste, a clinic directory, Google's review text and the ISHRS patient campaign. None of them is a secret. All of them take work.
Two things to know before you try it
The assistants disagree with each other. Perplexity named five clinics on the scarring question where ChatGPT named two, and it ranked them on review wording rather than protocol text. A practice can be the answer on one assistant and absent on another, and the fix is different for each.
And where you ask from changes the answer. We ran this from outside the Netherlands and both assistants still scoped to Amsterdam because Amsterdam was in the question, but on other runs we have watched the asker's location pull in businesses the question did not mention. Ask from where your customers are.
Find out which questions you already win
Send us your website and we run your customers' questions, generic and specific, across ChatGPT, Claude, Gemini and Perplexity, and send back which ones name you and which name someone else. Free, within a few working days. Or run it yourself with the Five Chats.
Van Found is an Amsterdam consultancy that gets clinics, law firms and other professional practices named when ChatGPT, Claude, Gemini and Perplexity answer. One client per category, at a time.