This is not a someday thing. It is already how a real slice of patients shop. In an October 2026 piece for Medical Economics, the numbers were blunt: about 70 percent of patients are open to or already using AI tools to research physicians, and 31 percent have used a generative tool like ChatGPT to look for a provider. Most striking, 26 percent said an AI recommendation directly influenced which provider they chose. For comparison, 28 percent cited a referral from their primary care doctor. The machine is now nearly as persuasive as the family doctor who has known you for years.
So the question is no longer whether patients ask AI for a doctor. They do. The question is how the AI answers, and what you can do to be the name it says.
The two kinds of AI searches, and only one is winnable
The single most useful idea we have seen on this comes from Carl White of MarketVisory Group, quoted in that same Medical Economics report. He splits what patients ask AI into two buckets.
The first is the teach me search. "Why does my heel hurt in the morning?" "Is Botox safe?" "What is TRT?" These are education questions. And here is the hard truth: a local practice almost never wins them. The answers get built from huge health publishers, medical associations, and national sites with enormous authority. As White put it, it is very difficult to win in the teach me category. Chasing it is a good way to burn time.
The second is the recommendation search. "Who is the best dermatologist near me?" "Find a pediatric dentist in Tampa that takes my plan." This is the one that turns into a patient. And this one is winnable, because the AI stops reaching for a textbook and starts reaching for local signals: who shows up strongly in the area, who has recent reviews, who clearly treats this exact thing in this exact city. That is your home field.
Where the AI actually gets your name
Here is the part that should calm you down: AI assistants are not pulling names out of a mysterious black box. They read the same public web that Google has read for years. When a patient asks for a doctor, the tool leans on a handful of sources you already influence:
- Your Google Business Profile. Your category, your service area, your hours, your reviews. This is the single biggest local signal, and it feeds both Google's own AI answers and, indirectly, the others.
- Your reviews, and how fresh they are. Volume, rating, and recency all tell the machine the same thing they tell a human: this place is real, busy, and trusted right now.
- Your website text. Not your logo, not your slider. The plain words that say what you treat, who you treat, and the town you sit in. A machine can only quote what you actually wrote.
- Trusted mentions elsewhere. Accurate listings in directories, local news, and health sites that line up with everything above.
Notice what is missing from that list: a secret setting, an insider contact, a credit card. You cannot buy your way into an AI recommendation the way you buy an ad. As the Medical Economics piece and our own work both keep showing, the practices that get named by AI are simply the ones that already did the local search homework. If you want the deeper version of this shift, we broke it down in why ChatGPT keeps recommending hospitals instead of your practice.
Why big systems win the lazy version, and how you beat them
Left alone, AI tools lean toward the biggest, most mentioned names, which is why a patient often hears the local hospital or a national chain first. That is the lazy answer, built on sheer size. But it is beatable, because size is not the same as relevance. A giant hospital system is a blurry match for "pediatric dentist who is good with scared four year olds in my zip code." A focused local practice that says exactly that, backed by fresh reviews from local parents, is a sharper match. The AI rewards the sharper match when the signals are there.
This is the same fight independent practices already run on Google, and it is winnable the same way. We wrote about the broader version in how an independent practice competes with hospital systems. The move is never to out shout them. It is to out specify them.
Reviews are doing double duty now
Reviews were always about trust. Now they are also raw material the AI reads to decide who to name. The Medical Economics report notes that 84 percent of patients check online reviews before choosing a provider and 51 percent read at least six. The machine is reading them too, and it treats a steady stream of recent, specific reviews as proof you are active and well regarded.
You do not need a mountain of them. Scott Bartnick of Otter Public Relations told Medical Economics that a practice needs only about 15 to 20 five star reviews to start gaining real traction. The catch is freshness and replies. A hundred reviews that stop in 2023 look like a closed business. Two new ones a week, each answered with a real human reply, look alive. Bartnick's point on replies is worth repeating: make them custom, not copy and pasted. A thoughtful response is a signal to the next patient and to the machine.
Write like you are answering the question
If there is one change that helps with AI the most, it is this. Stop writing website copy like a billboard and start writing it like an answer. A page that says "Compassionate care, redefined" tells a machine nothing. A page that says "We are a dermatology practice in Austin treating acne, eczema, and skin cancer screenings, accepting new patients and most PPO plans" can be quoted word for word when a patient asks.
That means plain, factual sentences. Your services named plainly. Your city and neighborhoods on the page. A real frequently asked questions section that answers the things patients actually type. This is the overlap where good healthcare SEO and AI visibility become the exact same job: you are feeding clear facts to anything that reads your site, whether that is a person, Google, or ChatGPT. The old idea that SEO and AI search are separate projects is already gone. It is one project now, and we cover the newer side of it in whether your practice should care about tools like Perplexity.
What this means for a practice owner this quarter
You do not need to chase every new AI tool or panic about the next one. The signals that win an AI recommendation are the same durable things that win local search: a complete and accurate Google profile, consistent details everywhere, a steady habit of fresh reviews with real replies, and a website that states plain facts about what you do and where. Do those, and you show up whether the patient uses Google, ChatGPT, Google's AI answers, or whatever lands next year. Skip them, and no clever trick saves you.
The practices that feel invisible to AI almost always have a gap in that foundation: a half filled profile, reviews that dried up, a website full of slogans and empty of facts. The good news is every one of those is fixable, and fixing them pays off in regular search at the same time.
How EtherealMinds sets this up
We work only with healthcare practices in the United States, so this is a daily conversation now. Owners ask us some version of "why does the clinic across town get recommended by ChatGPT and I do not?" The answer is almost never luck. It is that their local presence is tighter, their reviews are fresher, and their website speaks in facts a machine can repeat.
That is the work we do. We build the website that states your services and location in plain, quotable terms, we run the local and AI search work that earns the recommendation, and we tie it into the full patient acquisition system so a patient who hears your name from an AI can land on a strong page and book in a few taps. You cannot buy the recommendation. You can earn it, and that is exactly the part we handle.
Want to be the name AI recommends?
Book a free strategy call and we will check how your practice looks to Google and to AI tools right now, where patients are searching for what you offer, and what it would take to be the name that comes up when someone asks an assistant for a doctor near them.
Book a free strategy call →Frequently asked questions
How does AI decide which doctor to recommend?
AI assistants like ChatGPT and Google AI pull from the same public web that search engines read: your Google Business Profile, your website text, your reviews, and trusted directories and local pages that mention you. When someone asks for a doctor near them, the tool favors practices with a complete, consistent local presence, plenty of recent reviews, and clear plain text that states who you are, what you treat, and the city you serve. There is no secret button to push and no way to pay for the spot. The practices that already have strong local search signals are the ones that get named.
What is the difference between a teach me search and a recommendation search?
A teach me search is when a patient asks AI to explain something, like what causes heel pain or how Botox works. A recommendation search is when they ask the tool to name a provider, like best dermatologist near me who takes my insurance. As Carl White of MarketVisory Group put it in Medical Economics, it is very hard for a local practice to win the teach me questions, because big health publishers dominate them. The winnable game is the recommendation search, and that is decided by your local presence, your reviews, and clear information about your services and location.
Can I pay to have my practice recommended by ChatGPT or Google AI?
No. Unlike ads, there is no paid placement inside an AI recommendation. The assistant builds its answer from public information it trusts about you, so the work is earning those signals, not buying them. That means a complete and accurate Google Business Profile, consistent name, address and phone everywhere, a steady flow of recent reviews, and a website that states in plain words what you do and where. Ads still have their place for immediate visibility, but the AI recommendation itself is earned through your organic presence.
How many reviews do I need for AI to recommend my practice?
There is no exact cutoff, but you do not need thousands. Scott Bartnick of Otter Public Relations told Medical Economics that a practice needs roughly 15 to 20 five star reviews to start building real credibility. What matters just as much is recency and your replies. AI tools and patients both treat a steady trickle of fresh reviews as a sign the practice is active and trusted, while a wall of old reviews reads as stale. Ask a couple of happy patients every week and keep it going.
Is optimizing for AI search different from normal SEO?
It is mostly the same foundation with a few extra habits. AI tools read the same signals as Google, so a strong local presence, accurate listings, and good reviews help both. The extra work is writing in clear, factual, question and answer style so the tool can quote you easily, keeping your service and location details explicit rather than hidden in slogans, and making sure your site states facts a machine can lift. Good healthcare SEO and good AI visibility are now the same project, not two separate ones.