15 questions hair transplant patients ask AI, and what answers them
A hair transplant patient does not call five clinics anymore. They type 'hair transplant cost in delhi' into ChatGPT or Perplexity and trust whichever clinic the engine can name with confidence. I scanned 13 real Indian hair transplant clinic websites against the 15 questions patients actually type, and most of the gap between the clinics that get named and the ones that do not has nothing to do with surgical skill.
Hair transplant is a considered purchase. A patient researches for weeks before booking a consultation, and a growing share of that research now happens inside an AI answer engine instead of a search results page. ChatGPT, Perplexity and Gemini do not link to ten clinics and let the patient sort it out. They read a handful of pages, decide which ones actually answer the question, and name two or three clinics in a sentence.
So I ran 13 real Indian hair transplant clinic websites through the same deterministic engine that powers this site's free Report Card tool. I was not scoring how good the clinics are. I was scoring how readable each site is to the kind of machine that now sits between a patient and a consultation booking, the same check behind the question index at /ai-visibility/questions and the full writeup at the hair transplant AI visibility study.
The result surprised me. This is the strongest-scoring vertical I have checked so far. But it has one specific blind spot, and that blind spot happens to be the exact markup that answers the questions below: cost, grafts, timeline, pain. Here are the 15 questions, the fix for each, and what the scan found across all 13 sites.
The 15 questions, and the fix for each
Most clinics bury pricing inside a PDF brochure or a WhatsApp-only quote, which an answer engine cannot read or cite. Put an actual price range in visible text on the page, and mirror it in the offers.priceRange field of the clinic's schema. A number a machine can parse beats a number hidden behind a form.
If the price only exists in a PDF, it does not exist to the AI reading your page.
This is a yes-and-no answer that clinics usually explain somewhere in a long paragraph three screens down, if at all. Give it its own FAQPage entry with a direct first sentence: transplanted grafts are permanent, but non-transplanted native hair can still thin over time. A direct FAQ entry gets pulled into an answer verbatim; a buried paragraph does not.
The clearest one-sentence answer wins the citation, not the most thorough one.
Graft-count guidance almost always lives inside an infographic image, which is invisible text to a crawler. Rewrite the typical ranges (for example, a receding hairline commonly needs 1500 to 2500 grafts, stated as a range, not a promise) as real HTML text under a MedicalProcedure schema block describing the procedure type. Text an engine can quote; a picture it can only skip.
An infographic looks informative and reads as blank to a machine.
Patients want a comparison, and most sites either only offer one method or bury the comparison in marketing copy. Build an actual side-by-side as HTML text (not a scanned chart), with an answer-first opening line stating the core tradeoff: FUE leaves no linear scar and heals faster, FUT can move more grafts in a single session. Lead with the answer, then the detail.
A real head-to-head paragraph gets quoted; vague copy gets skipped.
Timeline questions need a month-by-month answer, not a single vague line like 'results vary.' Write it out as text: shedding phase around weeks 2 to 4, new growth starting around month 3 to 4, visible density by month 8 to 12. Put it inside an FAQPage entry so it can be lifted whole into an answer.
A timeline written as a sentence is machine-readable. A timeline shown only as a before/after slider is not.
This question is often answered honestly but only in a video testimonial, which text-based answer engines cannot transcribe or cite. Add a short, factual text answer next to the named surgeon's credentials, using a Physician schema entry so the answer is attached to a real, named practitioner and not an anonymous 'our team.'
A named surgeon answering in text carries more trust signal than an anonymous video.
Safety questions reward specificity, not reassurance. Describe the procedure honestly in text under a MedicalProcedure schema block: what the surgery involves, standard precautions, and who is not a good candidate. Vague safety badges as images do not carry any of that information to a crawler.
A badge image says 'safe.' A text paragraph says why.
Recovery timelines are a factual, low-risk answer to get right, and most clinics still leave it vague. State it plainly in an FAQ entry: most patients return to non-physical work within 3 to 5 days, avoid gyms for about 2 weeks. Precise day counts are exactly what an answer engine looks for to complete a direct question.
Vague reassurance loses to a specific day count every time.
This is the highest-trust question in the set, and it is answered with a named person, not a clinic logo. Publish the operating surgeon's name, qualifications and registration as visible text, wired into a Physician schema entry with medicalSpecialty set. A credentials PDF locked behind a download link does none of that work for an AI reading the page live.
E-E-A-T for a medical procedure means a named, credentialed human in the markup, not a logo.
Women's hair transplant is frequently folded into a single generic service page or a photo gallery with no text around it. Give it its own heading and an answer-first opening paragraph describing eligibility and approach for female pattern hair loss specifically. A dedicated H1 and a few sentences of real text tell the engine this clinic actually treats this case, not just files it under 'hair loss.'
A service buried inside a gallery does not get cited as a service.
Before/after galleries are almost always raw images with no alt text and no caption, so an answer engine can see that photos exist but cannot say what they show. Add real alt text and a caption to each image (grafts used, months post-op, technique) so the evidence is described in words, not just displayed.
A photo with no caption is invisible proof to a machine, even though a human can see it fine.
Reviews usually live as a screenshot of a Google listing or an embedded widget, neither of which a crawler can parse into a number. Add an aggregateRating schema entry tied to the clinic's actual review count and score, not a rounded or invented figure. A real, verifiable rating in structured data is what lets an engine cite 'rated 4.8 from 200 reviews' instead of saying nothing.
Only publish a rating you can back with a real review count. A made-up number is worse than no number.
Financing is often mentioned only in a sales call, never on the page. State the financing options in plain text (EMI tenures, any partner names, whether a down payment is required) as its own FAQ entry. A financing question answered in text can get surfaced directly; one that requires a phone call to answer cannot.
If the answer requires calling the clinic, the AI will answer with a competitor instead.
This is a local-intent, real-time question, and it needs structured hours, not a hand-written 'Mon-Sat, 10 to 7' line inside an image banner. Add an openingHoursSpecification block to the clinic's LocalBusiness schema, with the real address as text, not embedded only inside a map screenshot.
Open now is a schema field, not a banner graphic.
Beard and eyebrow transplants usually get one line at the bottom of the main hair transplant page, with no separate heading or price. Give the service its own MedicalProcedure entry with its own heading, its own price range, and a short answer-first paragraph on how graft count and technique differ from scalp work. A service without its own text is a service the engine cannot answer a question about.
If beard transplant only exists as a bullet point, it does not exist as an answerable question.
What I found scanning 13 hair transplant sites
What I found scanning 13 hair transplant sites: this is the strongest vertical I have checked so far. The median readability score was 89 out of 100, with a range from 56 to 100. Nine of the 13 scored an A, three scored a B, and one scored a C. Structured data (schema markup) was present on 12 of the 13 sites, missing on only one. Open-graph and social tags were present on 10 of 13. The one weak spot: only 3 of the 13 have a missing or broken main heading, so most sites at least get that basic structure right.
For comparison, when I ran the same scan across clinic sites more broadly the median score was around 77, and restaurant sites came in at a median of 79. Hair transplant clinics, at a median of 89, are the most machine-readable vertical I have scanned to date. But readability and answerability are not the same thing. Only 4 of the 13 sites have FAQ markup at all, which means 9 of 13 have none. That is the blind spot, and it happens to be exactly the markup built to answer the questions in this post: cost, grafts, timeline, pain. A site can score an A on general readability and still be silent on the one question a patient actually typed.
The full breakdown, clinic by clinic, is in the hair transplant AI visibility study. If your site is missing the FAQPage, MedicalProcedure or Physician schema this post keeps pointing at, the schema generator will build a valid block from your own facts in a couple of minutes, no invented numbers included.
The pattern under all 15
Do not try to fix all 15 questions this week. Read the list, find the one question closest to what your own patients actually ask in a consultation, and fix that single fact first: the price range, the graft-count estimate, the recovery days, whichever one is currently missing or buried in an image. One correct, machine-readable answer beats fifteen half-finished ones.
If you are not sure where your own site stands, run it through the free Report Card tool. It is the same deterministic engine behind every number in this post, it takes under a minute, and it will tell you exactly which of these gaps your site has before you guess.
Frequently asked questions
If I add FAQ schema, does that guarantee ChatGPT or Perplexity will name my clinic?
No, and I would not trust anyone who tells you otherwise. Schema makes an answer easier for the engine to find and quote, it does not force a citation. What it does is remove the excuse: if your page has a clear, direct answer in text and structured data, the engine can choose it. If the answer only exists in a PDF or an image, the engine cannot choose it no matter how good the clinic is.
Our before and after photos and price list are already on Instagram. Isn't that enough?
Instagram captions and photo posts are not something a website-reading answer engine can see as part of your site. The content needs to live as text on a page you control, with schema attached to it. Instagram is fine for social proof with humans, it just does not do any of this specific job.
We already have a good reputation and real patient results. Why does any of this matter?
It matters because a good reputation only helps you if the engine can read proof of it. A strong aggregateRating with a real review count, a named surgeon with credentials as text, and a captioned before/after photo all turn word-of-mouth trust into something a machine can verify and quote. Without that, the reputation is real but invisible to the tool a growing number of patients are now asking first.
Where does the schema code actually go, and do I need a developer?
Structured data is a small block of JSON that sits in the page's head section, and most website platforms let you paste it in without touching the rest of the site. The schema generator produces a valid block from your own facts, which you or a developer can drop in directly. It takes minutes, not a rebuild.