Is Google's AI Showing Patients Fewer Practices?

A dermatologist we know ran a little test on herself this summer. She opened Google on her phone, typed the kind of thing a patient would type, something like dermatologist near me for a mole check, and waited to see her own practice. It was there, technically. But she had to scroll past a full AI written answer first, one that had already picked three clinics, described what each was good at, and pretty much made the decision for a hurried reader before they touched a single listing. Two of those three were her competitors. Hers was further down, in the part most people never reach.

She called us a little rattled. Nothing about her practice had changed. Same reviews, same website, same ranking she had watched for years. What changed was the shape of the page. Google stopped handing patients a list and started handing them an answer. And an answer, by design, is short.

What actually changed on the search page

For twenty years, a local search worked like a directory. You typed something, Google showed a map with a pack of nearby practices, and you picked. The game was simple to understand even if it was hard to win: get into that pack of results and you got seen.

Now Google's AI often answers before the pack ever appears. It reads the reviews, the profiles, and the websites of the practices around you, then writes a tidy summary that recommends a handful and skips the rest. The patient reads a paragraph, not a list. And here is the uncomfortable part for practice owners: a paragraph has room for three names, maybe four. A map pack had room for more, and a full page of listings had room for plenty. The funnel got narrower at the exact top.

This is not a hunch. A 2025 study by the local search firm Whitespark looked at 540 local searches across several cities and industries, medical clinics included. AI Overviews, the AI written answer at the top, showed up on about 68 percent of local business searches. The traditional local map pack showed on only 39 percent. When someone asks an informational question, the kind patients ask constantly, like what does this symptom mean or which clinic treats this, the AI answer appeared over 90 percent of the time. The old blue links and the map are still there. They are just no longer the first thing, or sometimes the main thing, a patient sees.

Why fewer names means more pressure on you

Think about the math from the patient's side. In the old world, a patient searching for a physical therapist saw ten or fifteen options and browsed. Being the seventh best looking result still got you clicks, because people scrolled and compared. In the new world, the AI reads all fifteen and tells the patient about three. Being seventh does not get you mentioned. It gets you skipped, silently, before the patient even knows you exist.

So the question every practice owner should be asking is not where do I rank. It is am I one of the few the AI picks. That is a higher bar, and it rewards a specific kind of practice: the one with a clean, active, well reviewed presence that is easy for a machine to read and easy to feel good about recommending. AI does not want to send a patient to a clinic with a half empty profile and three reviews from 2021. It has a reputation to protect too.

There is a real risk hiding in here for practices that used to coast. For years, a lot of clinics ranked fine on the strength of being nearby and being old, even with a thin online presence. The AI layer is filtering those out. It is holding practices to the standard their rankings always implied but never really tested. If your online presence was weaker than your position suggested, this is the moment it shows.

What the AI is actually reading when it picks

Here is the good news buried in all this. The AI is not choosing at random, and it is not choosing based on who spends the most. It leans on the same signals that have always mattered for local visibility, just weighted harder and judged more strictly. Three things do most of the work.

Your Google Business Profile, and whether it looks alive. A complete profile with current hours, the right services and categories, real photos, and recent activity tells both the map and the AI that you are open, active, and worth recommending. Local SEO teams tracking rankings through 2026 keep reporting the same pattern: profiles that have not posted a photo, update, or post in the last month tend to slide. A profile that has gone silent reads as a practice that might have closed. Keeping it fresh is not busywork, it is a signal. We break down the mechanics in our guide to ranking higher on Google Maps for a medical practice, and it is the same foundation the AI reads from.

Your reviews, especially the recent, specific ones. This is the heaviest lever. A 2025 rater8 report covered by ConsumerAffairs found that 84 percent of patients check online reviews before choosing a new provider, 71 percent use reviews as their very first step, and 61 percent now trust reviews more than a recommendation from friends or family. AI tools pull from that same well of review text to decide who to name and what to say about them. A steady flow of recent, detailed reviews, ideally naming the provider and the service, is what gets a practice quoted in an AI answer. A stale set of old reviews gets it left out. If reviews are your weak spot, start with our piece on getting more Google reviews for a medical practice.

Your website, and whether a machine can read it. AI answers are built by software that scans your pages. If your site buries your services in an image, hides your location, or explains what you do in vague marketing language, the AI struggles to understand you and leans toward a competitor it understands better. Clear headings, plain answers to real patient questions, obvious services and locations, and a fast, mobile friendly build are what let AI describe you accurately and confidently. This is where good AI search optimization for a medical practice stops being a buzzword and starts being the difference between getting named and getting skipped.

The mistake practices are making right now

The wrong reaction to all this is to panic and chase some new AI trick, or to assume the game is rigged and give up on search. Both are wrong. There is no magic prompt that gets you into an AI answer, and search is not dead, it is just stricter.

The other mistake is subtler and more common: doing nothing because your rankings still look okay in the tools. Ranking well in a report that measures the old map pack can hide the fact that patients are now seeing an AI answer that never mentions you. You can be number four on paper and invisible in practice. The only way to really know is to search the way your patients do, on a phone, in your own town, and read what the AI says back. Most owners have never done it. Do it this week. It is often a wake up call.

How we think about it at EtherealMinds

We treat AI search as one more reason to get the fundamentals genuinely right instead of faking them. When we take on a practice, we get the Google Business Profile complete and active, build a real system for collecting recent reviews per provider, and rebuild the website so both patients and machines can understand it in seconds. That is the core of our healthcare SEO work, and it is deliberately boring in the best way, because boring fundamentals are exactly what the AI rewards.

It all lives inside the same patient acquisition system we run for practices across the country. Getting found is the first job, and the rules for getting found just changed under everyone's feet. The practices that win the next few years are not the ones chasing the newest hack. They are the ones with a presence so solid that whether a patient gets a list, a map, or an AI written answer, your name is in it. Fewer slots at the top only hurts if you were never truly earning your spot. If you were, it is a gift, because it clears out the clinics that were coasting.