AI is changing how patients choose doctors in 2026, with 47% of surveyed U.S. patients reporting that they have used AI to research medical providers, up from 31% in 2025. Patients are increasingly turning to platforms like ChatGPT, Google AI Overviews, and Perplexity for doctor recommendations. As AI influences provider selection, medical practices must prioritize accurate online information, strong patient reviews, and visibility in AI-generated answers.
In this video, Etna Interactive CEO, Ryan Miller breaks down new research on patient trust in AI search and what it means for medical practices. Ryan Miller explores why the accuracy of AI-generated information about your practice matters alongside your traditional search rankings.
Video Transcription:
Hey again, it’s Ryan Miller with Etna Interactive, and we’re here today to talk about AI search, how it’s evolved over calendar year 2026, and what some recent data says about consumer trust and use patterns inside of AI search systems has changed and evolved.
A year ago, we sent out our forecast talking about how fast we expected AI to influence consumers’ decisions. And it’s coming true, but in some surprising ways. Let’s dive into the data.
So in just the last couple months, there have been a couple reports that come out, the first of which directly from Google, that talked a little bit about changes in usage patterns. And specifically, they broke an important milestone with over a billion monthly users in their dedicated AI systems like Gemini.
But importantly, they also shared that about 2 and a half billion users every month were encountering AI overviews. Importantly, that’s now half of Google’s total usage is being either actively exposed to or is actively choosing Google AI results. Now, I think the bigger story is actually what’s happening outside of Google.
So, in the survey that came out from Rater8 just a couple of months ago, they took a look at a cohort of about 1,000 U.S. patients who they interviewed about their usage patterns over the last 12 months of AI systems. They’ve repeated this study now 3 times, and the evolution of the consumer behavior is really interesting.
And the first point that I think is really fascinating is how many providers have actively, or how many patients have actively used AI to help influence medical provider selection. The leap is big, 16 points, from 31% to 47%. Nearly half of everybody in the study cohort was now actively engaging with AI systems for provider selection.
Importantly, they dove deeper, asking questions about where they are most reliant, where they have the greatest trust for AI-based provider recommendations, or rather, for provider recommendations in general. And I thought this was fascinating, because for me, we are at a tipping point.
A lot like when, I don’t know, the thing that kind of comes to mind for me is the adoption of ride-sharing when compared to taxi use, and that moment in time when it flipped, and slightly more of the population was comfortable, you know, jumping in a stranger’s car than, you know, the preference to actually taking a taxi. And that moment has come and gone.
Within the last couple months, what the data shows is that when we look at a sample of patients, 32%, now, this is the smallest number, said that the thing they rely on most is a word-of-mouth referral. Slightly higher, 34% said, no, no, no, I’m still relying on a traditional Google recommendation, but this is the number that I think is most important in this study.
36% says they actually prefer freestanding AI search platforms like ChatGPT, Claude, Perplexity, and Grok. for a medical provider recommendation. They trust it more than Google, they trust it more than a word-of-mouth recommendation.
So, keep that in mind as we start to break down some more of the stats in that Rater8 report. Mature patients, in particular, they’re leading on AI adoption. And I’ll be honest on this one, this was not what I had predicted.
So, 64% of patients, age 45 to 60, put AI first, that it’s their preferred methodology for provider research. Only 28% among a younger patient audience. This is an important moment where you need to reflect on your patient demographic and decide what does this data point say about how important AEO, or Answer Engine Optimization, is for your practice.
Now, when we just look at the questions around how patients are perceiving Google’s first page, Google’s result set, which you know, remains, we’re talking about the #1 place where consumers are potentially going to be exposed to provider recommendations still today.
Ultimately, what we find is that, you know, as we work our way down the page, patients are trusting the AI summary the most, about twice as much as any other feature on the page.
Despite the fact that very often the AI result is just a synthesized recapitulation of the organic results that appear directly below. Now, importantly, when we look at that breakdown, what’s that delta? 37% of patients trusted the AI overview, the summary at the top of the page when it’s present, the most, compared to only approximately 20% in the organic result set, 17% in the local pack, and just, or 13% in the local pack, rather, and then just 7% in paid search results.
So, trust falls pretty quickly as we move through the different features on Google’s first page. But the takeaway here is AI overviews, like we talked about in last year’s forecast, are super important on Google’s first page.
Importantly, patients are trusting the AI-synthesized answer across all of the cohorts, even when the data is fundamentally wrong. The study allowed the testers to be able to analyze the quality of the answer set. Two-thirds of the patients in the study actually encountered incomplete information. It spanned a fairly wide range of inaccuracies that can emerge, from wrong specialty, incorrect hours, insurance or payment information being wrong, the wrong address, and, very often, linking a practice to a competitor’s phone number.
So, the interesting takeaway here is: People aren’t checking; they’re just taking at face value the information that’s coming back from AI systems today to the tune of about 60%. And that’s a risk. That’s a concern, a consideration that has definitely influenced all throughout calendar year 2026 the strategies that we deploy for our own clients.
There was one other data point that emerged from the Rater8 study that studies the links between AI visibility, consumer behaviors, and reputation that I think is worth surfacing here. And ultimately, there is a double whammy that can come with reputation if you’re not mastering it today.
So, one of the things that came out of the study was a big leap in the number of people who are walking away from clinics regardless of the previous exposure that they’ve had to them. Whether that be a word of mouth recommendation or using the clinic in the past, 55% are walking away. They will not go to a clinic when exposed to significantly negative reviews. That threshold was about 4.0.
Historically, we’ve studied, and we found like 4.2, but I’d say close enough. 75% will not book with a medical business that they encounter has an online reputation of 4.0 or below. And what’s important here, for those that have followed us and, you know, we talk about reputation a lot, we’ve studied it extensively, our own data suggests that practices are about 8 times more likely to get a review that relates to a customer service issue.
And that same finding was echoed in this most recent report from Rater8, with rude staff and not listening, patient not feeling heard, topping the list for things that are decrementing reputation and pushing patients away from a practice.
It’s not about clinical outcomes, which tend to be where especially surgical practice leaders will focus. It’s really about the service equation. Now, here’s the double whammy effect that’s potentially the risk here, is patients are being dissuaded from, you know, from reaching out to a practice when their reputation falls below a certain threshold.
But where do they go? They go to AI systems, for 39% of patients. AI systems to find an alternate provider. This signals 2 things that I think are important. One, it’s going back to the notion of how many people are starting to trust AI systems to guide them to the alternative.
But importantly, all AI systems are heavily influenced by reputation, and if your reputation is not stellar, it’s less and less likely that you’re actually going to be that alternate recommendation and benefit from AI systems being able to push people to you.
The bottom line: What does this all mean? So, AI is already answering for you. You can’t wait to ensure that it’s right. For Etna’s clients, all of our 2026 tactics have really focused on not only increasing the likelihood that you’re going to be cited by an AI system but increasing the accuracy and reliability of what’s said about you.
If you’ve been dragging your feet, or if your agency partner hasn’t been clear about exactly what you’re doing, it’s a great opportunity to make sure you dive head-in to a conversation about what you need to be doing today to ensure both your visibility and the accuracy of your practice in AI search.
Leave a Comment