Streben Research · A Guide for Practice Owners

How AI assistants actually recommend eye doctors

Patients are starting to ask ChatGPT which eye doctor to see, and it does not choose the way Google does. Here is what nearly 500 of our own tests showed about how it picks, and the short list of what moves it in your favor.

The short version

Patients are already asking AI which eye doctor to see, and its answer is built from different signals than a Google search.
For broad searches like "best eye doctor near me," your website is almost never the source. Your reviews, your Google Business Profile, and the directories AI trusts decide the answer.
For specific searches like "dry eye specialist" or "scleral lens fitting," your own service pages can win, if the page is clear and the AI can reach it.
Most "AI optimization" sold to practices is untested or invented. A short list of fundamentals is what works, and you should keep your normal SEO alongside it.

Patients are already asking AI for a doctor

This is not a someday problem. In a recent survey of nearly a thousand patients, 47% said they had used AI to find a new provider, up from 31% just months earlier. A separate consumer survey found 45% had used AI for local business recommendations, up from 6% a year before.

Two things make it easy to miss. Patients still verify before they book, usually by checking the reviews themselves. And the click almost never shows up in your website analytics, so AI can be steering patients toward you or past you without leaving a trace in your reports.

AI does not choose the way Google does

When someone asks ChatGPT or Gemini for an eye doctor, it does not read your website's marketing copy the way you might hope. It leans on structured, trusted records: your Google Business Profile and map listing, your reviews, and a small set of directories it already trusts. ChatGPT now reads a licensed Yelp feed directly. The practical point is simple: most of what moves an AI recommendation lives off your website.

What our research found

Because no one had cleanly measured this for local practices, we ran our own study: nearly 500 tests of ChatGPT across 11 US cities, asking the way real patients ask, and repeating each question many times because AI answers change from run to run. One pattern held up across every city: whether your website gets cited depends on how specific the question is.

of broad "best eye doctor" answers cited a practice's own website. Reviews and directories decided it.

9 in 10

of "dry eye specialist" answers cited a practice's own service page.

100%

of "scleral lens specialist" answers cited a practice's own service page.

In other words, the broad search most patients start with is won on reputation, not on your website. The specific searches, for a condition or a specialized service, are where your website earns its place, as long as the page clearly names what you do.

The twist worth knowing: the word "best"

We found one more lever. Adding "best" to a specific search flips it back toward reputation. "Dry eye specialist" cited practice websites nine times out of ten; "best dry eye specialist" cited them almost never, and ran on Reddit and review directories instead. A superlative is a request for a ranking, and AI answers a ranking from opinion sources. The more a patient asks for the "best," the more the answer runs on your reviews and reputation rather than your website, and that is exactly the highest-intent search of all.

What actually moves the needle

Stripped to what the evidence supports, the list is short, unglamorous, and mostly not about your website.

Earn reviews, steadily. Rating gets you through the gate (AI tends to recommend practices rated around 4.3 stars and up), and review volume is the strongest signal that separates the practices AI keeps naming from the ones it forgets. Highest leverage, and it lives off your website.
Complete and maintain your profiles. Google Business Profile, Yelp, Healthgrades, and Zocdoc are the records AI reads and cites. A stale or unclaimed listing is now a direct liability.
Make sure AI can actually reach your site. Some security setups quietly block AI search crawlers by default, which makes a practice invisible to AI while everything looks fine in your normal reports.
Write clear, specific pages for what you treat. Genuine, plain-language pages about your conditions and services are where your website does earn AI citations. Name the specifics: the equipment, the lens types, the conditions.
Keep those pages current, and measure honestly. Fresh content gets cited more often, and a single AI query is noise. Track how often AI names you across many checks over weeks, not one screenshot.

What to stop paying for, and what to keep

A lot of "AI optimization" is sold to practices right now: FAQ schema markup, special "llms.txt" files, and confident, precise-sounding "ranking factors." We reviewed the evidence, including the peer-reviewed research, and the honest finding is that they do close to nothing for AI recommendation, and some of the most-quoted statistics in the field are simply made up.

Important: keep your normal SEO

None of this means "SEO is dead." This is about the AI recommendation layer specifically. Classic Google and Bing search still reward good on-page work, structured data, and quality content, and ordinary search is still a large share of how patients find a practice. Keep your standard SEO for organic search, and add the off-site work above for AI. Two systems, and you want to show up in both.

Want your practice to be the one AI recommends?

This is exactly what our AI Visibility service does: we build and run the review engine, optimize your Google Business Profile, confirm the AI crawlers can reach your site, write the specific service pages that earn citations, and measure the whole thing honestly over time.

See the AI Visibility service Book a Free Growth Call