GEO for Medical Practices: How Clinics Get Named When Patients Ask AI for a Doctor

Patients ask the chatbot before they call the front desk

A patient's question to an AI engine is rarely 'who is the best doctor'. It is narrower and more practical, because the patient has a plan, a schedule and a symptom. Each question below has a factual answer that lives somewhere on the web. When that somewhere is a stale directory, the engine repeats stale facts.

• Best primary care doctor near me accepting new patients
• Dermatologist near me that takes my insurance plan
• Is this doctor board certified?
• Urgent care or the ER for a sprained ankle?
• Same-week appointment with a specialist in San Antonio
• Which pediatric practice near the Medical Center has Saturday hours?

Notice how many hinge on facts the practice controls: new-patient status, accepted plans, hours, credentials. Notice also the urgent care question. Engines answer general health questions too, and a practice that publishes careful, reviewed explanations has a chance of being the cited source rather than a stranger's forum post.

The public record an engine reads about a doctor

Medical answers draw on a mix of regulator data, health directories and general local sources. The table lists the ones with verified roles, what each holds and what a practice should do. Texas sources are shown because that is where we are based; other states have their own medical boards.

On health topics, Google says its systems give even more weight to content showing experience, expertise, authoritativeness and trust, because the stakes for the reader are high. That favors pages written or reviewed by named clinicians with real credentials, and it punishes generic health copy written to fill a blog calendar.

The doctor who left in 2023 is still taking appointments online

A common medical GEO problem is not invisibility. It is a practice described with last year's facts: a physician who retired still listed, an old suite number, a closed satellite office, a phone number that rings a fax. Engines lift these from directories and old profiles, and a patient who calls the wrong office does not try twice. The fix runs through entity optimization: one name, address, phone and provider roster everywhere.

Google's Business Profile rules are specific about practitioners. A doctor who is public-facing and can be reached at the verified location during stated hours may have a dedicated profile. When several practitioners share a location, the practice gets its own profile, separate from theirs, and each practitioner's profile carries only the practitioner's name. A solo physician at a branded practice shares one profile, named in the form 'Practice: Physician'. Departments with their own entrance and category, common in hospitals, can have their own profiles.

What Texas lets a practice say in an ad or on a web page

The Texas Medical Board's advertising rules reach websites, profiles and posts, not only paid ads. Rule 22 TAC 164.3 lists practices it treats as false, misleading or deceptive. Among them:

• A testimonial without a disclaimer or warning about the credentials of the person giving it and a disclosure of any compensation given for it.
• Offering a permanent cure for an incurable disease, sickness or illness.
• Claiming a unique or exclusive skill without substantiation.
• Failing to identify models or actors explicitly as models or actors rather than real patients.
• Using 'board eligible', 'board qualified' or similar wording calculated to suggest 'board certified'.
• Failing to disclose that an advertisement, article or infomercial is a paid presentation.

Rule 22 TAC 164.4 adds that a physician may advertise as 'board certified' only if currently certified by an ABMS member board, the AOA Bureau of Osteopathic Specialists, the American Board of Oral and Maxillofacial Surgery, or another board the Texas Medical Board has approved. Federally, the FTC's rule on reviews and testimonials (16 CFR Part 465) has banned fake or AI-written reviews, sentiment-conditioned incentives and undisclosed staff reviews since October 21, 2024.

Answering a one-star review without a HIPAA disclosure

Reviews feed AI answers, and replies feed them too. For a medical practice, the reply is also where the legal risk sits. Under HIPAA, a reply that reveals a patient's information is a disclosure. Trade press reports of HHS Office for Civil Rights actions describe providers penalized for exactly this: a Dallas dental practice ($10,000, 2019), New Vision Dental ($23,000, 2022), a North Carolina dental practice ($50,000 civil money penalty, 2022) and Manasa Health Center ($30,000, June 2023), which put mental health details in replies to Google reviews. HHS's own pages could not be checked, so treat those figures as reported, not confirmed.

Two points from American Dental Association guidance, written for dentists, carry over to physicians in our reading: a patient who names themself in a review has not waived their privacy, and even 'thank you for coming in' confirms that they are a patient. So the safe reply says nothing about the person, the visit or the care.

Getting new reviews has its own rules. Google bans incentives for reviews, asking only happy patients, and staff steering reviews toward specific content; asking every patient the same way is allowed. More on how engines use reviews is in reviews and AI recommendations, and AI quoting a bad review covers what to do when a complaint keeps surfacing.

Pages a patient, and a model, can lift intact

Practice websites tend to be either thin or vague: a stock photo, 'compassionate care' and a phone number. Engines cannot quote compassion. They can quote that Dr. A is accepting new patients, is board certified in dermatology by a named board, sees patients on Saturdays and is in network with a named list of plans. Each card below is a page or a block of facts worth publishing.

Structured data should mirror those pages. Schema.org has MedicalClinic and Physician types, plus IndividualPhysician and PhysiciansOffice beneath Physician, and properties such as medicalSpecialty, isAcceptingNewPatients, hospitalAffiliation and usNPI for the 10-digit National Provider Identifier. Google says no special markup is needed for AI search; the value is accuracy. Details are in structured data for AI. For voice assistants, see our guide to voice search for medical and legal practices.

A practice's first 30 days of GEO

1. 1

   Ask what patients ask

   Run six real patient questions through ChatGPT, Gemini and Google, plus our free AI visibility check, which queries engines through their APIs live. API answers can differ a little from the apps. Record every wrong fact and every competitor named.

2. 2

   Reconcile the roster

   List every physician, location and phone number. Compare against the medical board lookup, Google, Healthgrades, Yelp and Apple Business. Fix departed doctors and old suites first.

3. 3

   Set up practitioner profiles correctly

   Practice profile per location, practitioner profiles for public-facing physicians, following Google's naming rules.

4. 4

   Write the reply policy

   A short HIPAA-safe template for every review, approved by your compliance officer, and one person responsible for replies.

5. 5

   Publish the core pages

   Provider pages, insurance, hours and new-patient steps, each reviewed for 22 TAC 164 before it goes live.

6. 6

   Start a compliant review ask

   Every patient, the same message, no incentive, pointing to Google.

7. 7

   Re-run and compare

   Ask the same questions at day 30 and keep a monthly record. Some sources refresh slowly.

How DataCram works with a clinic

GEO is part of our monthly SEO and AI visibility programs: no long contract, a defined scope, from $625 a month, with most local businesses between $625 and $2,250, as set out on GEO pricing. The tier follows a free GEO audit. We track the questions patients ask across engines, clean up profiles and directories, draft HIPAA-safe replies for your approval, write provider and service pages for clinical and compliance review, and report each month what the engines actually said. AICiteBeam, the tracking platform we built, does the monitoring.

A solo practice with a full schedule and accurate listings may need none of this; fix the roster, write the reply template and stop there. Larger groups, specialists competing in a crowded area such as the Medical Center, and practices that keep being described wrong are where a program pays. Our healthcare industry page covers websites and marketing for practices, and the dental practice GEO guide covers the dental board's separate rules. Other regulated neighbors: law firms and accounting firms. Why regulated practices need sites that behave is the subject of this insight.

An engine cannot recommend a doctor whose own website never says she is taking new patients.