15 questions dermatology patients ask AI, and what answers them
A dermatology patient does not call five clinics and compare anymore. They open ChatGPT or Google's AI Overview, type the question they are actually worried about (acne scars, pigmentation, is this mole safe to ignore), and trust whichever clinic the answer names. I scanned the homepages of 17 real dermatology clinics across Mumbai, Delhi, Bangalore, and Hyderabad against the 15 questions patients type into AI most often, to see which ones are built to be named and which ones are invisible to the exact software their next patient is asking.
An AI answer engine does not skim a homepage the way a person does. It looks for structured, direct signals: a schema block that says this is a MedicalProcedure with this price, a FAQPage entry that states this fact in one sentence, a named physician with a stated qualification. A dermatology clinic can have a beautiful website and still be unreadable to the software increasingly deciding which clinic gets recommended.
So I scanned the homepages of 17 real dermatology clinics across Mumbai, Delhi, Bangalore, and Hyderabad, the same deterministic engine behind the free tools on this site, and checked each one against the 15 questions a dermatology patient actually types into ChatGPT, Perplexity, or Google's AI Overview. Not what the clinic wants to say about itself. What the patient is asking.
Below are those 15 questions, the specific fix (the exact schema type or on-page change) that lets a clinic answer each one in a way an AI can read and cite, and what the scan found across the 17 sites, including one gap present on 16 of them.
The 15 questions, and the fix for each
Most clinic pages talk about acne treatment in general paragraphs but never state what it costs, pushing the visitor to call or WhatsApp for a quote an AI engine cannot see. Add a MedicalProcedure schema block named 'Acne Treatment' with the offers.priceRange field filled in, and repeat that range in plain visible text on the page itself, not just in a downloadable brochure.
A price an AI cannot read is a price that does not exist to the AI.
Clinics that offer both laser resurfacing and subcision for acne scars often bury the distinction inside one long 'skin treatments' paragraph, which gives an AI engine nothing specific to cite for either procedure. Split them into two MedicalProcedure entries, each with its own heading, its own short description of who it suits, and its own priceRange, so a query about one is not answered by a page about the other.
One vague paragraph cannot answer two different questions. Give each procedure its own heading and its own schema entry.
Pigmentation and melasma are usually mentioned once in a list of conditions treated with no real explanation, so there is no answer for an AI to lift out. Write a short FAQPage entry that states what causes melasma, what the clinic actually does about it (peels, topical regimens, laser), and roughly how long a visible result takes, in plain sentences an engine can quote directly.
A condition named in a list is not the same as a condition answered in a sentence.
This is a trust question, and trust signals need to be machine readable, not just a photo and a name on the About page. Mark up the treating dermatologist with Physician schema including their MD or DVD qualification and medicalSpecialty, and attach an aggregateRating field tied to a real, current review count rather than leaving reputation to a separate Google listing the AI may never connect to the page.
If your dermatologist's credentials only exist in a photo caption, they do not exist to the AI deciding who to recommend.
PRP for hair fall is one of the most searched procedures on a derm site, yet many clinics describe it only in a hair loss video or an Instagram reel, formats an answer engine cannot parse. Add a MedicalProcedure entry named for PRP hair treatment, state the typical number of sessions and the priceRange in visible text, and let the schema mirror exactly what is on the page.
A video explains PRP to a person. A schema block explains it to the AI recommending a clinic.
This is a question worth answering honestly rather than avoiding: patients are going to ask it whether a clinic addresses it or not. A factual FAQPage entry that explains what brightening or pigmentation correcting treatments can realistically achieve, that results vary by skin type and are not permanent 'whitening', and that a dermatologist consultation should set expectations before any procedure, gives the AI an accurate, responsible answer to cite instead of leaving the question to less careful sources.
An honest answer to an awkward question is still an answer an AI can cite. Silence is not.
Psoriasis and eczema are chronic conditions patients research heavily, but most clinic pages fold them into a generic skin conditions line with no detail on approach or duration. A short FAQPage answer covering how the clinic manages flare ups versus long term control, in plain text rather than a linked PDF guide, gives an AI engine an actual passage to quote when someone asks which dermatologist handles chronic skin conditions.
Chronic conditions need a real paragraph, not a bullet point in a services list.
Anti aging procedures are usually the most price sensitive query on a derm site, and also the most likely to be quoted only in person. Add MedicalProcedure schema for botox and for fillers separately, each with an offers.priceRange stated per unit or per area in visible text, since a single combined range for both procedures gives an AI nothing precise enough to cite.
Combine two procedures into one vague price range and you have effectively priced neither for the AI reading your page.
Chemical peels are cheap enough that patients expect to compare clinics on this question specifically, but many pages mention peels only as a bullet under facials and skin treatments. Give the peel its own heading, its own short paragraph on what it treats and how many sessions it typically takes, and its own MedicalProcedure entry so the answer engine has a dedicated passage rather than one buried word.
A treatment worth one line in a bulleted list is a treatment the AI will skip past.
Safety and pain questions need a direct, quotable answer, not reassurance buried in a testimonial. A FAQPage entry stating that mole and wart removal is typically done under local anaesthesia, roughly how long recovery takes, and the priceRange for the procedure gives the AI a factual passage instead of forcing it to guess or stay silent.
Reassurance in a testimonial is not an answer. A direct sentence about pain and cost is.
This is a common enough query that a clinic answering it well, honestly, without dismissing home care, becomes the source an AI cites for the more serious when to see a dermatologist half of the question. A short FAQPage entry naming the signs that dandruff needs a dermatologist (persistent flaking, scalp irritation, hair thinning) rather than a shampoo change gives the engine something concrete to quote.
Answer the home remedy question honestly and you earn the right to be cited on the harder one.
Laser hair removal pricing varies by body area and session count, and clinics that leave this to a phone call lose the query entirely to a competitor who states it. Add a MedicalProcedure entry per major area (face, underarms, full body) with its own offers.priceRange in visible text, so the AI can match a specific query like underarm laser cost to a specific, citable number rather than an average buried in a brochure.
One average price for every body area answers no specific question well.
Side effect and safety questions are exactly what FAQPage schema was built for: a direct question, a direct answer. State plainly what the common, temporary side effects of a peel are (redness, mild peeling) and who should avoid it, in a real FAQ block rather than a disclaimer PDF a patient has to download and an AI cannot read at all.
A safety answer locked in a downloadable PDF is a safety answer the AI never sees.
A consultation fee is the smallest number on the page and often the one clinics are most reluctant to publish, but it is also the single most asked question before a first visit. State the fee in visible text, add LocalBusiness schema with an openingHoursSpecification so the AI can also answer when are they open, and keep both the fee and the hours current, since a stale number is worse than none.
The smallest price on your site is often the first thing a patient asks an AI about. Publish it.
Before and after photos prove results to a human eye but say nothing to an AI engine, which cannot see the image, only read around it. Write a real caption under each photo describing the procedure, the number of sessions, and the approximate timeframe, and use descriptive alt text rather than a filename like IMG_2041, so the result itself becomes something the AI can cite.
A photo an AI cannot read is not proof to the AI. A caption is.
What I found scanning 17 dermatology sites
Across the 17 clinics, the median readability score was 89 out of 100, with a wide spread from 29 to 93. Grade wise, 11 scored an A, 2 a B, 3 a C, and 1 an F. Structured data, the schema markup that tells an AI engine what a page actually contains, was present on 13 of the 17 and missing on 4.
The 4 clinics without structured data had a median score of just 57, against 90 for the 13 that had it, a 33 point gap from one missing category of markup. But the sharper gap is FAQ markup specifically: only 1 of the 17 clinics had it. That is the exact schema type that answers a direct question, like most of the 15 above, in a form an AI can quote. 16 of 17 clinics are running a derm practice with real answers to real patient questions and no machine readable way to say so.
If your clinic is one of the 16, the fix is not a redesign. It is adding FAQPage and MedicalProcedure schema for the specific questions your patients actually ask, which the free schema generator on this site will build for you without needing a developer.
The pattern under all 15
Do not try to fix all 15 questions at once. Pick the one closest to what your own patients actually call and ask about, acne cost, PRP for hair fall, whatever it is for your practice, and fix that page first. A clinic that answers three questions well and machine readably will get cited more often than one that vaguely covers fifteen.
Then run your own site through the free Report Card to see your actual score and which of these gaps apply to you, the same scan used for the 17 clinics above.
Frequently asked questions
does chatgpt actually recommend specific dermatology clinics
Increasingly yes, when it has something specific to cite. AI answer engines pull from pages that state facts directly, like a procedure, a price range, or a named physician's credentials, rather than pages that only describe a clinic in general marketing language. A clinic with no structured data or FAQ content is much harder for an engine to name with confidence.
what is faqpage schema and do I need a developer to add it
FAQPage schema is a small block of structured data that pairs a question with its answer in a format search and AI engines can read directly, the same block that was present on only 1 of the 17 clinics scanned. You do not need a developer to add it: a schema generator tool can build the code from your existing content and you paste it into the page.
why does structured data matter more than good writing
Good writing still matters for the person reading the page, but structured data is what tells an AI engine what the writing actually means, this paragraph is a price, this one is a procedure, this one is a physician's qualification. The 33 point score gap between clinics with and without structured data in this scan reflects that difference directly.
how often should a dermatology clinic re-scan its site
Every time you add or change a procedure, price, or physician on the site, since the schema needs to match what is actually on the page. Beyond that, a re-scan every few months is enough to catch drift, a price that changed on the page but not in the schema is worse than no schema at all.