Search for "dermatologist" and AI hands you a clean shortlist: Curology, Nurx, the direct-to-consumer skincare names everyone in the category recognizes. Add one word, your city, and the answer falls apart. Ask for a dermatologist in Atlanta or Dallas or Boston, and the model names dozens of practices and crowns none of them. The national race has clear frontrunners. The local race has no leader at all.
That gap is the opportunity. Bain & Company found that 80% of consumers now rely on AI-written results for at least 40% of their searches. Google itself reports that AI Overviews are driving a 10%+ increase in search usage for the types of queries that trigger them. In healthcare, patients are already acting on it. A 2026 rater8 survey reported by TechTarget found that 36% of patients said AI swayed their choice of doctor, outpacing Google search results for the first time. Most practice owners read "AI won't rank us" as a wall. It's an open door.
Why dermatology AI search has no leader
Dermatology is one of the most fragmented categories in local healthcare, and fragmentation is exactly what makes it winnable. According to IBISWorld and Stifel research, the top four industry players account for only about 2% of total market share, and about a third of dermatology practices are solo operations. In our analysis of 27,812 AI answers to 6,953 consumer-health prompts across ChatGPT, Gemini, Perplexity, and Claude, more than 7,200 dermatology practices surfaced in the answers we tracked. No metro leader cleared 30% share of those local answers. We break down the methodology in our guide to AI visibility metrics.
Compare that to the national picture. The same research found the top three direct-to-consumer names cluster within four points of each other: Curology at 35%, Miiskin at 33%, Nurx at 32%. Nationally, a few brands own the conversation. Locally, no one owns the conversation.
"Fragmented" is not an abstraction here. It means no single practice has accumulated enough citations — instances where an AI model names or links to a specific source in its answer — in any given city to become the default answer. The concentration swings hard by metro. Atlanta is the most consolidated market we measured, where one practice (Olansky) holds about 46% of local answers. Boston is the most open, where the leader sits near 14%. A practice in Boston competes against a field where everyone is small. A practice in Atlanta has to unseat an incumbent. Either way, the marginal page moves the needle, because no one has locked the category.
Fragmentation also shows up across assistants: for the same dermatology prompt, our four AI engines agreed on only about 7% of the practices they named. There is no single AI answer to win, which is another reason to own your city on your own pages rather than chase any one engine.
AI search in dermatology is not impossible to win. It is unclaimed.
The condition-and-city page is the unit of victory
If you build one thing, build condition-and-city dermatology pages: a single page for each condition you treat in each metro you serve. Not a national "Our Services" page. Not a catch-all "Conditions We Treat" list. One page for "acne dermatologist in Atlanta." One for "Mohs surgery in Dallas." One for "eczema treatment in Boston."
AI assembles each local dermatology answer from a mix of source types: community forums like Reddit, reference sites like Wikipedia, government and academic authorities, third-party directories like Healthgrades, and providers' own websites. In our analysis, provider sites account for about 61% of cited links across the full market, but that number is spread across more than 32,000 domains. The largest single practice site, US Dermatology Partners, captured only about 0.5% of all citations. Your own site is the one layer in that mix you fully control, and most practices badly under-use it. That is what condition-and-city pages fix.
The template is simple. Each page covers one condition in one metro, written for the patient actually searching:
What the condition is and how your practice diagnoses it, in plain language.
The specific treatments you offer for it, named.
Who treats it at your practice, with credentials.
Local detail: the metro, neighborhoods served, what to expect at a first visit.
The instinct is to consolidate, to write one strong acne page and serve every city from it. That instinct loses. A model answering "acne dermatologist in Atlanta" matches the page that says exactly that. Specificity wins.
If you run multiple locations, the architecture for scaling these pages without creating thin duplicates deserves its own treatment. We cover that in our guide to generative engine optimization (GEO) for healthcare clinics.
Medical versus cosmetic: two different citation games
Dermatology runs on two revenue lines, and they earn citations in different ways. Treat them the same and you'll underperform in both.
Medical pages compete on clinical authority. Acne, eczema, psoriasis, skin-cancer screening: these are health questions, and AI models weight clinical trust signals heavily when answering them. A medical page earns its place by reading like medicine. Name the board-certified physician who treats the condition. Describe the diagnostic process. Reference the treatment protocols. Skin disease affects roughly one in four Americans, per the American Academy of Dermatology, so the search volume is real and the patient intent is serious. Pages that demonstrate genuine clinical depth get cited as sources; pages that read like brochures get skipped.
Cosmetic pages play a different game. For medical dermatology, the gatekeeper sources AI reads are clinical: Healthgrades and Yelp surface most, with Healthgrades the single most-cited domain at 2.2%. For cosmetic work, review platforms carry more weight, because a patient choosing a Botox provider is closer to choosing a restaurant than choosing an oncologist, and the AI answer reflects that. Cosmetic pages need volume and recency of reviews, before-and-after specifics, and pricing transparency where you can offer it. Clinical authority still helps, but it doesn't carry a cosmetic page the way it carries a medical one.
The practical takeaway: grade your medical and cosmetic pages on different criteria. Don't pour clinical credentials into a Botox page and expect them to do the work reviews do. Don't thin out a skin-cancer page with promotional language meant for the aesthetics lane.
The trust signals AI models read
Models can't interview your physicians, so they read proxies for trust. Three matter most, and most practices underuse all three.
Board certification is the strongest single credential signal. The American Board of Dermatology is explicit that being "certified" is not the same as being "licensed": certification attests to specialized training and ongoing assessment that a license alone does not. State it plainly on every provider bio and every condition page. "Board-certified dermatologist" is not marketing language to a model; it's a verifiable credential it can attribute.
Provider bios are the second signal, and they should read like evidence, not filler. Name the physician, the certification, the medical school and residency, the conditions they treat, and the procedures they perform. A condition page that ties the treatment to a specific named, credentialed provider gives a model a clean fact to cite.
Schema markup is the third, and it's the one most practices skip. Think of schema as a set of labels you add to your website's code. A human visitor never sees them, but they tell search engines and AI systems exactly what your page contains your practice name, address, provider credentials, specialties in a standardized format the machines read directly. Yolando's free Chrome extension can grade any page for AI readiness, including schema gaps. Two schema types matter most for dermatology practices. MedicalClinic labels your location pages with details like address, hours, and accepted insurance. Physician labels your provider pages with fields for medical specialty, board certification, and hospital affiliation. Fill those fields in, and you give AI models a clean, machine-readable summary of exactly why your practice is qualified. Your web developer or SEO agency can implement both; the Schema.org documentation linked above walks through every available field.
The absence of these signals costs you. In our research, the single most-cited domain in local dermatology answers was Healthgrades, at 2.2% of all cited links, more than four times the largest individual practice site. A directory out-cites real practices because it publishes structured, credential-rich, consistently formatted provider data at scale. Your own pages can match that level of data quality, and you have the advantage of being the primary source for your own providers. But you have to actually build the pages and add the markup. Most practices haven't.
Own your city's answer
National dermatology brands have the lead. Your city does not have one yet. Publish specific, credential-backed condition-and-city pages. Mark them up so models can parse the trust signals. Score medical and cosmetic pages on separate criteria. That is how a practice becomes the name AI returns first.
Yolando maps which condition-and-city pages your group can realistically own, builds the citation strategy behind them, and tracks where you're gaining ground against directories and incumbents. See how Yolando can grow your clinic's AI visibility.
