15 questions fertility patients ask AI, and what answers them
A fertility patient no longer scrolls ten blue links. They ask an answer engine the heavy questions first, success rates, cost, IVF versus IUI, and it names a clinic or two in a sentence. Here are 15 questions patients actually ask, and the exact thing your site needs so the clinic it names is yours.
Before a fertility patient calls a clinic, she asks an AI. "IVF success rate after 35." "Cost of one IVF cycle in Bangalore." "IVF vs IUI for PCOS." I keep seeing questions like these typed into ChatGPT and Gemini, and the answer engine has to pick a clinic to name or describe based only on what it can actually read on a page.
A clinic with real, careful outcomes and a genuinely good doctor can still get skipped. Not because the care is worse, but because the success rate lives in a brochure PDF the engine cannot parse, the doctor's name appears with no credential attached to it anywhere in the HTML, and the FAQ is a screenshot dropped into the page as an image. The engine cannot cite what it cannot read.
Below are 15 questions patients actually type, and for each one, the specific on-site fix that turns unreadable good work into something an answer engine can find, trust and quote. None of these fixes are marketing. They are structure: named authors, dated pages, honest numbers presented as text, and schema that states clearly what a page is and who stands behind it.
The 15 questions, and the fix for each
These are the kinds of prompts patients type into an answer engine, in the spirit of the question index and grounded in the clinic study. For each one, the fix is not to promise more. It is to move the answer from a place a machine cannot read, and cannot attribute, to a place it can.
MedicalWebPage schema and a datePublished field so the engine can see the data is current and attributable.FAQPage schema around a "how much does IVF cost" question so the range can surface directly in the answer engine's response.FAQPage schema so it can be lifted whole.MedicalClinic schema with your address, phone, hours and service list, and make sure your name, address and phone number match exactly across your site, your Google Business Profile and any directory listing. Mismatched NAP details are one of the quietest ways a genuinely local clinic gets left out of a local answer.Physician or Person schema entry with their qualification, registration number and years of practice, and put that named doctor's byline above a plain-text explanation of how endometriosis affects IVF planning. An unnamed "our team" page reads as nobody to an engine trying to attribute a medical claim.FAQPage entry, citing general clinical guidance on cumulative success across cycles rather than a specific promised number, and note that the right answer depends on individual factors your doctor should discuss. A dated, named, sourced answer beats a vague "every case is different" with nothing behind it.MedicalWebPage schema, a named author, and a last-reviewed date so the engine treats it as current.MedicalWebPage schema. Name the counsellor or coordinator who handles donor cases so the page has a real, attributable author.HowTo or clear numbered headings, covering stimulation, retrieval, transfer and the wait. An engine can lift a numbered, dated, plain-text sequence far more easily than a rotating homepage graphic.FAQPage entry that says a second opinion is a normal, sensible step, and briefly explains what a second opinion review usually looks at. This kind of candor is also a strong general trust signal to an engine weighing which clinic to name.MedicalWebPage schema and a reviewedBy field.FAQPage schema.FAQPage schema with a Physician author. A clinic that hands patients the right questions in advance reads as the one with nothing to hide.What the scan found, and why it matters more here
None of this is theoretical. In a scan of 16 clinic websites, including fertility clinics, the median site scored 77 out of 100 on a plain readability and structure test, and not one of the 16 scored an A.
The clearer split was around structured data. Sites with no structured data at all had a median score of 54. Sites that had it scored a median of 80.5. That is not a small gap, and it lines up with what these 15 questions show: an answer engine cannot separate a good clinic from a badly documented one until the site tells it, in code and in plain text, what is true and who said so.
On a health topic like fertility, that structure carries more weight than usual. An engine, and a patient, is right to be suspicious of a number with no source. Named authorship, a review date, and honest data presented as text instead of a PDF are not decoration. They are what makes a real answer possible.
The full breakdown is in the clinic study, and the same pattern shows up for dental patients too. If writing JSON-LD by hand sounds like a chore, the schema generator reads your page and drafts a correct MedicalClinic or FAQ block for you to paste.
The pattern under all 15
If you run a fertility clinic, you do not need to rewrite your whole site to start showing up in these answers. Pick the three questions above closest to what patients actually ask you, and fix the on-site structure behind them first: a named doctor with real credentials, one honest sourced table instead of a PDF, and an FAQPage block around your most common question.
My free Report Card scores any clinic homepage out of 100 in about ten seconds. It will not diagnose your medicine, but it will tell you exactly where the structure is missing, so an answer engine finally has something real to read.
Frequently asked questions
Will fixing my website actually change what ChatGPT says about my clinic?
It is not guaranteed, but it meaningfully improves your odds. Answer engines pull from what they can read and attribute, so a page with a named doctor, a dated honest answer and clean schema is far more likely to get cited than the same information locked in a PDF or an image. Think of it as removing obstacles, not buying a result.
Is it safe to publish our real success rates online?
Yes, if you present them honestly: the actual numbers, the year, the sample size, and who reported them. Patients and answer engines both distrust a bare percentage with no source, so an honest table with context tends to build more trust than a rounded, unsourced headline number. Never publish a rate you cannot stand behind if a patient asks about it in person.
We do not have a dedicated writer. Who should be the named author on these pages?
It should be whoever actually knows the content: the treating doctor, embryologist or clinic coordinator, with their real name and qualification attached. A generic "clinic team" byline reads as nobody to an engine trying to attribute a medical claim. It does not need to be long, just real and consistent across pages.
How often should we update the success rate and cost pages?
Review them at least once a year, or sooner if your numbers materially change, and update the date shown on the page each time you do. A dated, recently reviewed page signals that the information is current, which matters more on a health topic than almost anywhere else on your site.