Local AI visibility: how your practice wins in its own catchment area

Why the search for a doctor is fundamentally changing right now

Just two years ago your potential patient typed "dermatologist Stuttgart appointment" into Google and scrolled through a list. Today she asks the same question in full sentences to ChatGPT, or she gets an AI summary from Google right at the top, before every classic result. The machine delivers not ten blue links but a recommendation: a name, an address, a sentence about the practice. Whoever is named there wins the inquiry.

For doctors this is a break with everything that ran for years under search engine optimization. It is no longer only about ranking on page one but about appearing in the one generated answer. Generative Engine Optimization, GEO for short, describes exactly this discipline: the targeted work on getting AI systems to know your practice, describe it correctly and recommend it at the right moment. For a location-bound practice this is not a gimmick but the new foundation of patient acquisition.

Important to understand: the AI does not invent its answers. It draws its knowledge from what appears online about you, your website, your Google Business Profile, review portals like Jameda or Doctolib and regional directories. If these sources are thin, contradictory or outdated, the machine produces either nothing or something wrong about you.

What patients really ask the AI

People type differently to AI assistants than to Google. Instead of keywords they ask whole, very concrete questions. Typical for the medical field: "Which pediatrician in Cologne-Ehrenfeld is currently taking new patients?", "Is there a GP near me with evening consultation hours?" or "Where can I get a short-notice appointment with an orthopedist with statutory health-insurance approval in Freiburg?". These questions contain location, need and often a constraint.

Exactly these details are your opportunity. A practice whose online presence clearly states that it takes new statutorily insured patients, offers an evening consultation and is located in a certain district gives the AI the building blocks it needs for a precise recommendation. A practice that only has an address and a phone number online stays invisible for such questions, even if it is perhaps professionally the best in the area.

Do the test yourself: open ChatGPT or Gemini and ask the five questions with which your desired patients would look for you. Are you named? Is the information correct? Often the result is sobering, wrong opening hours, an old address or no mention at all. This small exercise is the most honest starting point for everything that follows.

The Google Business Profile is your most important signal

For local visibility there is no more important single source than your Google Business Profile, formerly Google My Business. AI systems, above all AI Overviews and Gemini, draw on it massively. A fully completed profile with the correct specialty, exact consultation hours, services, photos and directions is the basis. If something is missing here or contradicts your website, the AI becomes uncertain and leaves you out when in doubt.

Pay special attention to the categories. Do not just enter "doctor" but the precise specialty, for example "specialist in dermatology" plus fitting additional categories like "skin cancer screening" or "allergist". These terms are the vocabulary with which the AI assigns questions to your services. Also maintain the attributes such as barrier-free access, parking or digital appointment booking, exactly what patients filter for in their questions.

Consistency is decisive. Name, address and phone number must be written identically everywhere online, on the website, at Jameda, in directories. Experts call this NAP consistency. Even a deviating spelling of the street or an old mobile number creates contradictions that an AI interprets as uncertainty and that lower your chance of a recommendation.

Your website must be readable for machines

Many practice websites consist of beautiful pictures and terse sentences like "A warm welcome". For people that seems friendly, for an AI it is empty. Language models need clear, factual text: which services do you offer, for which complaints are you responsible, which insurances, which consultation hours, which district, whether you take new patients. Write these facts out instead of hiding them in graphics or PDFs that machines read poorly.

Structured data helps additionally. With the schema markup "MedicalClinic" or "Physician" you can store specialty, opening hours and address machine-readably. This sounds technical but is a manageable task for your web maintainer and a strong signal that tells the AI unmistakably who you are and what you do.

Think in real patient questions. A dedicated subpage that answers the question "What to do about persistent back pain?" in a professionally sound way and at the end points to your consultation is exactly the material from which AI answers are built. This way you combine professional authority with local findability, without lapsing into dubious promises of cure.

Reviews and mentions are your reputation currency

AI systems assess not only whether you exist but also how you are talked about. Numerous, current and differentiated reviews on Google, Jameda or Doctolib are strong trust signals. It is less about the pure star count than about the substance: if patients name concrete things in reviews, short waiting time, empathetic explanation, good accessibility, that gives the AI exactly the keywords it picks up in recommendations.

Actively but properly ask your satisfied patients for a review, for example with a small card note at reception or a link after the appointment. Pay strict attention to the professional and data-protection limits: no bought reviews, no incentives, no disclosing patient data. Respond to criticism factually and without ever naming treatment details, your handling of reviews too is read along by the AI and by patients.

Mentions beyond the review portals are also valuable: an article in the local newspaper, an entry in your city's health portal, being named in the network of your medical association or a physician network. Such independent sources increase the probability that the AI classifies you as an established, real practice and does not skip you.

The medical special case: trust and regulations

Health is a sensitive topic for AI providers. Systems like ChatGPT and Google are deliberately cautious here and rely more strongly on reliable, verifiable sources. For reputable practices this is good news: whoever provides correct, consistent and professionally sound information is preferred, while blaring advertising rather hurts. Restraint and precision pay off doubly here.

Observe the legal framework. The Medicines Advertising Act, the medical code of professional conduct and the GDPR set narrow limits. No misleading promises of success, no before-and-after images where impermissible, no patient stories without express consent. Good GEO work for doctors never means getting louder but informing more clearly and completely than the competition.

Precisely because AI systems filter strictly in the health field, an advantage arises for those who do it right. A practice with a maintained profile, a fact-rich website and genuine, properly solicited reviews is classified by the machine as trustworthy, and exactly these practices land in the recommendations when a neighbor asks for a good doctor.

How to proceed concretely

Start with an honest inventory. Ask the typical patient questions to ChatGPT, Gemini and in Google search with AI Overview and note where you are missing or misrepresented. In parallel, check your Google Business Profile and all portals for completeness and identical information. These gaps are your priority list.

Then work systematically: complete the profile, align the NAP information everywhere, supplement the website with real, spelled-out facts and a few professional question-and-answer pages, have schema markup built in and establish a calm, permanent process for reviews. None of this is a one-off project, consultation hours, services and team change, and your online presence has to follow.

Measure success in the right place. Casually ask new patients at reception how they came to you, and repeat the AI tests every few weeks. If your practice increasingly appears there correctly and prominently, that is no coincidence but the result of this work, and the most direct way to steer the inquiries from your own catchment area to you instead of to the competition.

Your 90-day roadmap for the practice

Start small but committed. In the first thirty days you get your Google Business Profile in shape: correct opening hours, services with the right terms (not just "internal medicine" but also "travel medicine" or "ultrasound"), photos from the waiting room and reception. In parallel, check whether name, address and phone number are written exactly the same on the website, Jameda and Doctolib. This consistency is the foundation on which every AI answer rests.

In days thirty to sixty you take care of your website. Have structured data built in for a medical practice and write one clear page per treatment focus. From day sixty you establish a fixed review process: every medical assistant knows when and how to ask satisfied patients for honest feedback. After ninety days you have not a campaign but a routine, and exactly that is what the systems reward.

An example from everyday practice

Imagine a GP group practice on the edge of town. New patients used to find it through the neighbor's recommendation. Today someone types into the AI assistant: "GP nearby who takes new patients and has evening consultation hours." If the profile contains no hint of free capacity or evening appointments, the practice simply does not appear in this answer, even though it would fit perfectly.

The solution was unspectacular: the team added the note "New patients welcome" to the profile, entered the Thursday evening consultation as its own time and answered three common questions directly on the website. Four weeks later the first calls came with the sentence "The AI suggested you to me." Visibility here arises not through advertising but through clean, verifiable information.

Where local AI visibility reaches its limits

Be honest with yourself: AI systems replace neither good medicine nor a good reputation. They amplify what is there anyway. If patients leave the reception dissatisfied, no matter how well-maintained the profile is, it does not help. The technology makes you findable, but it cannot optimize away poor accessibility or long waiting times. So expect no miracles but a fair chance to be seen.

Then there are the rules of your profession. Sensational promises of cure or invented awards catch up with you faster in the AI world than before, because contradictions between sources become visible. Stick to verifiable facts. And reckon with the systems changing monthly, what works today you will want to check again in half a year.

Common questions from the practice

"Do I now have to become a technology expert myself?" No. You have to deliver the right information and know what matters. The technical building-in of structured data is handled by your web agency, the profile fine-tuning often by a trained staff member on the team. Your task is that the content stays medically correct and current.

"What if competitors do the same?" Then details decide: genuine reviews, complete service descriptions and fast accessibility. This is exactly where diligence pays off. Whoever consistently maintains the basics wins in their own catchment area, not the one with the biggest budget but the one with the cleanest signals.