How Patients Actually Choose a Clinic in ChatGPT vs Google

Quick answer: In Google, a patient chooses a clinic by scanning ten links and comparing. In ChatGPT, a patient gets one synthesized answer — a recommendation assembled from a small set of sources — and the clinic that isn't in that answer never enters the decision at all. The difference isn't the interface. It's that Google shows you a market, and ChatGPT shows you a verdict. The patient doesn't see the clinics that were omitted, and you never learn you were one of them.

• Google presents options; ChatGPT selects one. The patient's job changes from comparing to confirming
• AI answers are built from a narrow, repeatable source pool — registries, societies, directories, institutional pages — not from a general web crawl of every clinic site
• In "AI's Sources of Truth: How Chatbots Cite Health Information" (July 2025; 5,472 citations across ChatGPT, Gemini, Claude, Perplexity), 62.4% of all citations came from domains rated DR 81–100, with 30.7% from health media and 22.9% from academic sources (study summary)
• A clinic can be excellent and still absent. Being omitted is invisible to both the patient and the clinic — that is the structural problem
• The clinics that win in ChatGPT are the ones whose external sources agree about them, not the ones with the best website alone

Google: the patient compares

In Google, the patient types "best IVF clinic near me" and gets a page of results. They read titles, skim reviews, check two or three clinic sites, maybe call the first one that replies. The patient is doing the work of comparing. Your clinic can win that comparison with a good site, good reviews, and a fast phone pickup.

The key property: you are in the search results. Even if you rank eighth, you exist in the consideration set. The patient decides whether to include you.

ChatGPT: the patient confirms

In ChatGPT, the same question returns one answer: "Based on the available information, the highest-rated fertility clinics in your area include X, Y and Z." That is a shortlist, not a market. The patient is confirming a recommendation, not building a comparison set.

The shortlist is assembled from what the model's sources say — and those sources are mostly not your website. Published citation research on AI health answers shows the pool is concentrated: institutional and high-authority domains dominate, and the same source families appear again and again. The seven sources AI reads for fertility maps exactly which ones they are.

Two properties follow.

First, concentration. When a patient asks a general question, AI recommends a handful of clinics — not all the viable ones. In the October 2026 fertility benchmark, the top four clinics captured 36.6%, 32.6%, 23.7% and 22.6% valid recommendation coverage respectively, and the rest of the field sat in single digits or lower. The full table is here. The long tail barely exists in AI answers.

Second, invisibility of omission. If your clinic isn't named, the patient never sees it. There is no page 2 to scroll to. You don't lose the comparison — you are removed from the category. And the patient has no idea you exist, so no one tells you.

What determines who gets named

The same forces that move the benchmark: what the sources AI reads say about you. Registry data (CDC ART, SART) that matches your site's numbers. Directory records (FertilityIQ, Healthgrades, WebMD, Google Business Profile) that are complete and consistent. Answer-shaped content that matches how patients phrase questions. When these agree, AI can name you confidently. When they conflict, engines omit rather than adjudicate — a site that leads with one success-rate definition while SART reports another is the cleanest example of a conflict that gets a clinic dropped from answers.

The first 30% of a page matters disproportionately — in an analysis of 18,012 verified ChatGPT citations, 44.2% came from the first 30% of a document (Search Engine Land, 2025). Answer first, explain after.

What this means for your clinic

• You now have two separate visibility problems. Ranking on Google and being named in ChatGPT run on different systems that read different sources. Winning one doesn't win the other
• The comparison page is gone. You can't out-convert a patient who never sees you. The work is being in the answer, not being better in the comparison
• The patient's question is the entry point. The clinics named are the ones the sources match to the exact phrasing — "highest success rates," "best IVF clinic near me," "cost of IVF in [city]." Write your pages in the shape of the questions

What to do this week

• Run the 20-minute check on ChatGPT, Gemini, Claude and Perplexity with your city and your procedure — the exact prompts are here
• Compare the answers to your Google ranking. If you rank top-3 on Google but are absent from AI answers, you now know exactly which problem you have
• Ask what the AI answer says about you, not just whether you appear. Accuracy of the description drives whether you stay on the shortlist next month

One honest caveat

Search behavior varies by patient and by engine. The citation percentages above come from published studies at specific points in time; engines and their source pools change, and the October 2026 fertility benchmark is a snapshot, not a prophecy. This describes mechanism, not any individual clinic's quality. None of this is medical advice.

Frequently Asked Questions

Don't patients still use Google?

Yes — most patients use both, often in sequence: a general question first, then a deeper check. The shift is that the AI answer now happens first and narrows the set dramatically. The clinic omitted from the AI answer never makes it to the Google comparison.

Isn't ChatGPT just summarizing search results anyway?

Not exactly. ChatGPT's answers are synthesized from its training data and retrieved sources, weighted by authority and consistency — not a simple copy of a Google results page. That's why the source pool is so concentrated and why the two visibility problems differ.

Which matters more: ranking in Google or being named in ChatGPT?

They serve different moments. Google captures the patient who is still comparing; ChatGPT captures the patient who is ready to shortlist. For a clinic with limited marketing budget, being named in AI answers is the newer, less contested, and often cheaper win — but you should want both.

If I'm already top of Google, am I automatically in ChatGPT answers?

No. The two systems read different sources and weight them differently. A clinic can hold the top organic position for "IVF [city]" and still be absent from what ChatGPT says — the source pool decides.

How do I check what AI says about my clinic without paying anyone?

The same 20-minute walkthrough works for any practice: ask each engine the questions your patients ask, with your city and procedure in them, and note who's named. How to check AI visibility has the full method.

Curious what ChatGPT says about your clinic today? Claim Free Snapshot — I'll run the check and tell you where you appear, before anyone talks retainers, or see how the audit works first.

In Google you compete for clicks. In ChatGPT you compete for being named. Make sure you're in the answer.