Almost no patient asks AI for a clinic by name. They ask about the discomfort, the procedure or who accepts their insurance near home, and the clinic only makes it into the answer if the assistant finds a source that connects its name to that question. That is why how a clinic shows up in AI answers depends on three things the clinic itself can organize. The first is one page for each procedure, signed by the practitioner who performs it. The second is the same address, hours and insurance data on the website, the Google Business Profile and directories. The third is mentions of the practitioner in third-party sources.
This guide is for owners and managers of medical, dental and aesthetic clinics. The concept of GEO, optimization for AI-generated answers, is covered in the post what is GEO; here the subject is the clinic, with the questions patients ask, the sources the assistant reads and the health advertising rules that also apply to this channel.
What do patients ask ChatGPT before booking an appointment?
A health question asked to an assistant tends to be long, in the patient's own words and with context. The table shows five formats, with examples of the kind we list when building a clinic's question sheet, and what each one requires from the website.
| Patient question | What it is looking for | What the clinic needs to have published |
|---|---|---|
| "Which doctor should I see for knee pain when climbing stairs?" | The specialty, based on a symptom. | A specialty page stating which complaints it evaluates, signed by the doctor who sees patients. |
| "Does laser teeth whitening make teeth sensitive?" | Information about a procedure. | A procedure page covering what it is, the aftercare and the dentist's name with their CRO number. |
| "Gynecologist who takes my health plan in Tatuapé, São Paulo" | Practitioner, insurance and neighborhood in the same sentence. | Insurance plans written out as text, by specialty, and an address with the same neighborhood as the Google profile. |
| "Pelvic physical therapy near Santana subway station with appointments after 6 p.m." | Neighborhood, access and hours. | Actual hours on the website and the profile, with the nearest station or landmark. |
| "Psychologist or psychiatrist: who should I see first during an anxiety attack?" | A comparison between two specialties. | Informational text that explains the difference without advising on an individual case, reviewed by the person who signs it. |
The first two usually get an explanatory answer, often with no clinic cited. The last three call for a place, and that is where the assistant starts listing names. In practice, the clinic is competing for two different spots: being the source of the explanation and being the address recommended for care.
How a clinic shows up in AI answers: the sources the assistant reads
The assistant answers with what it learned in training or with what it searches on the web at the time of the question. In health, the sources it finds fall into four groups:
- The clinic's website, as long as the assistants' bots can open the pages. A block on the server removes the website from ChatGPT search without any warning, and the technical check for this is covered in how to show up in ChatGPT as a source.
- The Google Business Profile, with category, address, hours and reviews. Search AI features draw on Google's index, according to the AI features and your website documentation, and the profile is one of the few places in that index the clinic can edit on its own.
- Booking platforms, insurers' lists of in-network providers and practitioner directories, which show insurance, neighborhood and rating in the same format for dozens of clinics and are therefore handy for comparison.
- Institutional sources, such as hospitals, specialty societies and public agencies, which usually support the explanatory part of the answer. The clinic does not replace these sources, but it can appear alongside them when the procedure page is specific and signed.
The practitioner's name runs through all four groups. A page without the practitioner's name, registration number and review date is text without an author, and in health the assistant has plenty of sources with authors to choose instead.
What should the clinic publish on its own website?
The website is the only source the clinic fully controls. For it to serve the questions in the table, each item below needs to be in text, on a page the bot can read:
- One page per procedure and one per specialty, with the name of the practitioner who performs it, their CRM or CRO number and, for medical specialists, their RQE, the specialist qualification registration.
- Accepted insurance plans written out as text and broken down by specialty, not in a banner of logos, which is not always read as a list.
- An address with the neighborhood and a landmark, and hours identical to those on the Google Business Profile, including holidays.
- The name and registration of the medical director, which CFM Resolution 2,336/2023 requires in advertising for medical facilities.
- Common questions about each procedure answered right in the first sentence, with no promise of results and no advice on an individual case.
- The date of the last review and structured data describing the clinic and the practitioner, in the format explained in Google's introduction to structured data.
Profile, directories and insurers: what to check beyond the website
Beyond the website, the work is making every source tell the same story about the clinic. We start with the Google Business Profile: the right categories for each specialty, reviewed hours and replies to reviews that do not confirm the person is a patient. Care with health data in reviews and the split of what the clinic approves are detailed on the page on SEO for clinics and health and beauty businesses.
Next come booking platforms and each insurer's list of in-network providers. In the checks we run, old addresses and practitioners who have already left the clinic often show up in these listings, and the assistant repeats the wrong data with the same confidence as the right data. Finally come mentions of the practitioner as a source in news stories and health websites, which reinforce the authorship of the website's pages.
What is outside the clinic's control?
Knowing how a clinic shows up in AI answers includes knowing what no clinic, and no vendor, can control:
- The assistant chooses its sources for each question, and the same question can bring a different list the next day or in another app.
- For symptom questions, assistants tend to answer in general terms and recommend an appointment, without naming a clinic.
- The location of the person asking and the insurance plan they mention change the names that appear.
- Third-party listings, such as an insurer's provider list, can only be corrected by whoever maintains them, and the correction takes as long as that company takes.
In organic search, Busca Orgânica works with a contractual performance guarantee, with goals agreed at the start of the project. For a clinic, presence in AI answers is not part of those goals: it appears in the report as a reading of the tested questions, never as a commitment, and what the contract covers is in what goes into a guaranteed SEO contract.
Health advertising: the mistakes that keep a clinic out of AI
The mistakes below apply to medicine, dentistry and aesthetics. In dentistry, the reference is the Dental Code of Ethics and the CFO advertising rules; in medicine, the CFM resolution cited above.
- Before-and-after photos as bait. Outside the conditions the board's rule allows, the image creates ethical risk and explains nothing the assistant can cite.
- Promising results. Phrases like "painless" or "guaranteed results" are prohibited in medical advertising and make the page look like an ad, not a source.
- Price as bait. A time-limited discount on a procedure attracts the wrong question; before publishing any price, the clinic should check its own board's rules.
- AI-written text published without review. Google does not prohibit using the tool, but it requires accuracy and usefulness in its guidance on AI-generated content, and in health a technical error signed by a doctor becomes a professional board issue.
- AI bots blocked as a precaution. The clinic closes its website to assistants and then wonders why they only cite the booking platform.
- Mixed-up locations. Two or three locations with the same phone number and the same text keep the assistant from recommending the right location for the asker's neighborhood.
How Busca Orgânica handles this with clinics
At Busca Orgânica, SEO, GEO and AEO are part of the same project, because all three fronts touch the same clinic pages. With our own search engine optimization methodology, we start with the specialty's question sheet and with who the assistants cite today; only then do we organize pages, profile and listings. The full design of the service is on the page on GEO to get your brand cited by AI. Throughout the process of understanding how a clinic shows up in AI answers and acting on it, the person who reviews and signs the clinical content is the clinic's practitioner, never our team.
If you want to see what ChatGPT and Gemini answer today when a patient asks about your specialty in your neighborhood, talk to Busca Orgânica and request a review of your clinic's questions.