Vertical Market Evolution
The Six-Screen Search for a Dentist
Patients now spread their trust across an average of six discovery sources, including AI answers, while most dental practices still can't see what those sources are saying about them.
Part of Vertical Playbooks in the Insights library.
Abstract
The evidence points to something most dental practices haven't priced in yet: Google's share of local discovery is slipping, review count and star rating are only part of what earns trust, and AI tools went from a rounding error to a real recommendation channel for local businesses in a single year. Meanwhile, dentists are adopting that same technology fast in the back office and almost nowhere patients can see or hear it, and independent testing shows the AI tools patients are starting to trust are still inconsistent. Nobody has yet counted, specifically for dentistry, what those AI answers actually say about any given practice, which is the open question this study leaves standing.
U.S. dental spending reached $189 billion in 2024, up 4% after inflation over 2023, with government dental programs growing fastest, according to the ADA Health Policy Institute's analysis of federal health expenditure data. That growth is landing on a discovery surface that looks nothing like it did even two years ago. Among general local-business consumers (the best evidence available, since a dental-specific version of this research doesn't yet exist), BrightLocal's 2026 Local Consumer Review Survey found Google's share as a review or recommendation source fell from 83% to 71% in a single year, while use of AI tools such as ChatGPT for local recommendations jumped from 6% to 45%, now the third most-used source behind Google and Facebook. Apple Maps usage nearly doubled, and reliance on local news as a discovery source fell. The average consumer now checks six different sources before choosing a local business. None of this is dental-specific yet, and that gap matters, but there is no reason to expect dental search behaves differently, and the direction of travel is unmistakable.
That fragmentation is changing what consumers trust and how they decide, not just where they look. BrightLocal found 40% of consumers trust AI platforms for business recommendations against 32% who don't, 42% trust AI recommendations as much as traditional reviews, and 82% read the review summaries an AI engine writes when deciding on a business, with 23% willing to rely on that summary alone. At the same time, a British Dental Journal analysis of nearly 49,000 online reviews of NHS dental practices found that staff professionalism and communication, not clinical outcomes, were the biggest drivers of both satisfaction and dissatisfaction in what patients actually wrote. And when researchers tested AI chatbots directly on dental questions, the answers were inconsistent: one study found no chatbot was reliably both accurate and empathetic, and a separate study of teenage-anesthesia decisions in Turkey found that when parents turned to social media for advice, it measurably reduced anxiety and raised confidence, but 60% were also exposed to negative content along the way. Trust is moving toward AI-mediated answers faster than the accuracy of those answers has caught up.
For a practice trying to figure out where to put its effort, the picture is a market where attention has already spread past the one channel most owners still watch, and where a real, peer-reviewed framework shows that what a source contains, its statistics, citations, and specificity, can change whether a generative engine cites it by as much as 40% in general-purpose testing. What doesn't exist yet is a dental-specific count of how often ChatGPT, Google's AI Overviews, or Perplexity actually surface or recommend a given practice, and on what basis. That is the genuine frontier, not a settled fact: the terrain has moved, the tools to read it precisely for a given dental market are still being built, and a practice that starts mapping where its patients' attention already sits, rather than assuming it's still all in one place, is working from the more current picture.
The data, in one read
A Market That Keeps Growing, No Matter How Patients Find Their Way In
U.S. dental spending hit $189 billion in 2024, about 3.6% of all health expenditure, and grew 4% after inflation over 2023, roughly $7 billion in new spending, according to the ADA Health Policy Institute's analysis of CMS data. The growth wasn't spread evenly: government dental program spending rose 9%, largely reflecting Medicare expansion, while out-of-pocket spending grew 3.3% and private-insurance dental spending grew 2.3%.
That means more people are becoming dental consumers for the first time, or returning to care they'd deferred, right as the tools they use to find a dentist are being rewritten underneath them. This figure is U.S.-only; no comparable verified global number sits in the evidence for this study, so the growth story here is read as a U.S. one.
The Discovery Surface Is Fragmenting Under Practices' Feet
BrightLocal's 16th annual Local Consumer Review Survey, covering general local-business consumers rather than dental patients specifically, found Google's share as a review or recommendation source fell from 83% to 71% between 2025 and 2026, while use of AI tools for local recommendations jumped from 6% to 45% over the same year, now the third most-used source behind Google and Facebook. Apple Maps usage nearly doubled, from 14% to 27%, while reliance on local news sites as a discovery source fell from 48% to 29%.
Put together, consumers now check an average of six different review or recommendation sources before deciding on a local business. This is general-population research, not a dental-specific study, and that limit is worth holding onto. But there's no evidence dental search runs on a separate track, and a practice watching only its Google listing is watching a shrinking share of a wider picture.
In a single year, AI tools went from a 6% niche to a 45% mainstream channel for local recommendations, and Google's share slipped from 83% to 71%. That's not a small shift. That's the map changing while most businesses are still looking at the old one.
Patients Already Trust What the AI Answer Tells Them
The same BrightLocal survey found 40% of consumers trust AI platforms for business recommendations, against 32% who don't, and 42% say they trust AI recommendations about as much as they trust traditional reviews. Even more strikingly, 82% read the review summaries an AI engine writes when deciding on a local business, and 23% say they'd be willing to rely on that summary alone, with no further checking.
That's not a distant hypothetical. It's already how a meaningful share of consumers decide. The trust is arriving ahead of any independent measurement of whether it's warranted, which is exactly the gap this study keeps returning to.
What a Review Is Actually Measuring
A sentiment and topic analysis of 48,862 online reviews of NHS dental practices in England, published in the British Dental Journal, found consistently high positive sentiment (81 to 83%) across the periods before, during, and after COVID-19, with negative reviews falling from 13.87% to 9.72%. The researchers' key finding: staff professionalism and communication, not clinical outcomes, were the primary drivers of both satisfaction and dissatisfaction in what patients actually wrote.
A separate, single-source claim deserves more caution. A dental marketing agency, RevUp Dental, reports in its own call-tracking data that practices with fewer than 100 Google reviews typically generate under 200 new-patient calls a month, while practices with 500 or more reviews generate 400 to 600-plus, describing a 'tipping point' near a 4.7-star average. The sample size and method behind that figure aren't disclosed, so it stays flagged as contested: a real, published claim from a single vendor, not independently verified evidence to build a strategy on by itself.
When You Actually Test the AI Answer, It's Inconsistent
A study in the Journal of Prosthetic Dentistry tested five AI chatbots (ChatGPT-3.5, DeepSeek R1, Claude 3.5 Sonnet, Google Gemini, and Microsoft Copilot) on geriatric denture-related patient questions. Google Gemini scored highest on accuracy (mean 3.3, plus or minus 0.50) and Microsoft Copilot scored lowest (2.5, plus or minus 0.58, statistically significant). Copilot scored highest on empathy while ChatGPT-3.5 scored lowest, and accuracy and empathy were negatively correlated across the five tools. No chatbot tested was reliably both accurate and warm.
A separate study in BMC Oral Health compared ChatGPT-4 and Gemini against the World Dental Federation's own answers to its frequently asked oral-health questions. Both chatbots performed comparably to the Federation's answers on completeness and clarity, with ChatGPT-4 ahead of Gemini on the study's accuracy measure and closer to the Federation's answers on relevance overall; both were judged a workable, dependable general information source.
Patients are already leaning on these tools for real health information, and the tools themselves are still uneven. That matters because whatever an AI engine says when a patient asks a dental question, or asks about a specific practice, is being taken largely at face value.
Practices Are Moving Fast in the Back Office, and Barely at All Where Patients Can See
The ADA Health Policy Institute's dentist panel survey found 43.3% of U.S. dentists report using AI for at least one task in their practice, with another 26.4% planning to. That adoption concentrates almost entirely on non-clinical, administrative work: imaging and diagnostics (22.8%), insurance verification (13.6%), business analytics (10.1%), and front-desk check-in (10.1%). Under 5% use AI for patient treatment recommendations.
The same survey found active resistance further out: 82.6% of dentists do not plan to use AI for treatment recommendations, and 68.3% do not plan to use it to explain clinical findings to patients, citing fear of misdiagnosis or overtreatment and concern about eroding human clinical judgment.
That caution is reasonable on its own terms. But it's a decision about what practices do with AI internally. It says nothing about what AI engines are already saying about those same practices externally, unprompted, when a patient searches. That's a different question, and right now it's mostly going unanswered.
Dentists are being careful about AI in the exam room. Fair enough. But patients are already asking AI tools about dentists outside the exam room, and almost nobody in the practice is watching what gets said back.
A Very Analog Problem Still Sits Underneath the Digital Story
A narrative review in the Journal of Patient Experience, built around Andersen's Behavioral Model of Health Services Use, found that 14% of patients avoid dental visits altogether because of scheduling conflicts with work hours, and 46% report difficulty booking an appointment because of busy phone lines. The review frames digital or online booking as a genuine access lever at the pre-visit stage of the patient journey.
This matters because it sits right at the handoff. Winning attention across six discovery sources and earning a favorable AI-mediated answer counts for little if the practice on the other end is still hard to reach by phone. The bottleneck isn't always visibility. Sometimes it's what happens the moment someone tries to act on it.
The Open Frontier: Nobody Has Counted This for Dentistry Yet
A peer-reviewed framework presented at KDD 2024 showed that specific content-level changes, adding statistics, citations, and quotations to a source, can raise that source's visibility or citation share inside generative engine answers by up to 40%, tested against a purpose-built benchmark of real user queries. That's solid, general-purpose evidence that what a source contains changes whether an AI engine cites it. It has not been tested on dental content specifically.
And here is the limit of everything above it: no independent, dental-specific census exists yet of how often ChatGPT, Google's AI Overviews, or Perplexity actually surface or recommend a given dental practice, or on what basis they do it. Every AI-trust and AI-usage figure in this study comes from general local-business research, not dentistry. That gap is real, and it's why actually capturing what these engines say about specific practices, methodically and repeatably, is the next step for any practice or market that wants to know where it actually stands rather than assume it.
We know AI answers change what patients see. We know patients already trust those answers. What we don't yet have, for dentistry specifically, is a count of what those answers are actually saying about any given practice. That's the measurement gap worth closing.
The evidence, in numbers
Key findings, dated and sourced
-
U.S. national dental expenditure reached $189 billion in 2024 (3.6% of total health expenditure), a 4% inflation-adjusted increase over 2023. Growth was driven mainly by a 9% rise in government-program dental spending (Medicare expansion), plus 3.3% growth in out-of-pocket and 2.3% in private-insurance dental spending.
established American Dental Association (ADA) Health Policy Institute, ADA Health Policy Institute analysis of CMS National Health Expenditure data (2026-01)
-
43.3% of U.S. dentists report using AI for at least one task in their practice as of mid-2026, and another 26.4% plan to; adoption concentrates on non-clinical, administrative work (imaging/diagnostics 22.8%, insurance verification 13.6%, business analytics 10.1%, front-desk check-in 10.1%), while under 5% use AI for patient treatment recommendations.
established American Dental Association (ADA) Health Policy Institute, HPI Perspectives dentist panel survey (2026-07)
-
Dentists remain resistant to AI on clinical and patient-facing tasks: 82.6% do not plan to use AI for treatment recommendations and 68.3% do not plan to use it to explain clinical findings to patients, citing fear of misdiagnosis and overtreatment and erosion of human clinical judgment.
established American Dental Association (ADA) Health Policy Institute, Same ADA HPI dentist panel survey (2026-07)
-
Among general (non-dental-specific) local-business consumers, Google's share as a review or recommendation source fell from 83% in 2025 to 71% in 2026, while use of ChatGPT and other generative AI tools for local business recommendations jumped from 6% to 45% year over year, making AI the third most-used recommendation source behind Google and Facebook.
emerging BrightLocal, Local Consumer Review Survey 2026 (16th annual edition) (2026-02-11)
-
40% of consumers say they trust AI platforms for business recommendations versus 32% who do not, and 42% trust AI recommendations as much as traditional reviews. 82% read the review summaries an AI engine writes when deciding on a business, and 23% say they would be willing to rely on that summary alone.
emerging BrightLocal, Local Consumer Review Survey 2026 (2026-02-11)
-
Consumers now use an average of six different review sites when researching a local business in 2026, reflecting fragmentation of the discovery surface beyond Google (Apple Maps usage nearly doubled from 14% to 27% year over year; local-news-site usage as a recommendation source fell from 48% to 29%).
emerging BrightLocal, Local Consumer Review Survey 2026 (2026-02-11)
-
A peer-reviewed Generative Engine Optimization (GEO) framework, tested on a purpose-built benchmark of real user queries and sources, showed that content-level interventions (adding statistics, citations, quotations) can boost a source's visibility and citation share in generative AI engine responses by up to 40%.
established Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan, Deshpande, "GEO: Generative Engine Optimization" (KDD 2024)
-
A narrative review of dental patient experience, organized around Andersen's Behavioral Model of Health Services Use, found 14% of patients avoid dental visits due to scheduling conflicts with work hours and 46% report difficulty booking appointments because of busy phone lines, framing digital or online booking as an access lever for the pre-visit stage of the patient journey.
emerging Seyed Kian Haji Seyed Javadi, Shahid Beheshti University of Medical Sciences, "From Patient Experience to Dental Service Return Visits: A Narrative Review of Strategies for Improving Dental Care Services," Journal of Patient Experience (2025-12-09)
-
A sentiment and topic analysis of 48,862 online reviews of NHS dental practices in England found consistently high positive sentiment (81 to 83%) across pre-, during-, and post-COVID-19 periods, with negative reviews falling from 13.87% pre-pandemic to 9.72% post-pandemic; staff professionalism and communication, not clinical outcomes, were the primary drivers of both satisfaction and dissatisfaction in the review text.
established Ali Feizollah and Matthew Byrne, University of Manchester, "Patient perceptions of English dental services before, during and after the COVID-19 pandemic: a sentiment and topic analysis of online reviews," British Dental Journal (2025-12-12)
-
Testing five AI chatbots (ChatGPT-3.5, DeepSeek R1, Claude 3.5 Sonnet, Google Gemini, Microsoft Copilot) on geriatric denture-related patient questions found Google Gemini scored highest on accuracy (mean 3.3 plus or minus 0.50) and Microsoft Copilot lowest (2.5 plus or minus 0.58, P<.001); Copilot scored highest on empathy while ChatGPT-3.5 scored lowest, and accuracy and empathy were negatively correlated across chatbots (r=-0.152, P=.016). No chatbot was reliably both accurate and empathetic.
established Sivakumar I, Arunachalam S, Gadde P, Sharan J, "Performance of AI chatbots in responding to geriatric patient questions on denture issues: A mixed method study of accuracy and empathy," Journal of Prosthetic Dentistry, 135(5):e170-e177 (2026-05)
-
Comparing ChatGPT-4 and Gemini answers to the World Dental Federation's (FDI) frequently asked oral-health questions, both chatbots performed comparably to FDI's own answers on completeness and clarity; ChatGPT-4 outperformed Gemini on the study's accuracy criterion and was more similar to FDI's answers on relevance. Both were judged a prevalent and dependable general oral-health information source.
emerging Arpaci A, Ozturk AU, Okur I, Sadry S, BMC Oral Health (2025-08-02)
-
In a cross-sectional survey of 385 parents in Turkey, 80% searched social media for information on general anesthesia and sedation before their children's dental treatment, and 64.6% said that information influenced their treatment choice. Social media use was associated with a statistically significant drop in reported anxiety (7.18 to 6.09, p<0.001) and rise in confidence (5.31 to 6.82, p<0.001), but 60.1% of participants were also exposed to negative content, rising to 66.7% among heavy users (3+ hours a day).
established Fidancioglu YD, Catak T, "The role of social media in parents' approaches to dental treatment procedures under general anesthesia and sedation: a cross-sectional survey in Turkey," BMC Oral Health (2026)
-
A dental-marketing agency's proprietary call-tracking dataset reports practices with fewer than 100 Google reviews typically generate under 200 new-patient calls a month, while practices with 500 or more reviews generate 400 to 600-plus calls a month, with a reported tipping point around a 4.7-star average rating. Sample size and methodology are not disclosed.
contested RevUp Dental (dental marketing agency), RevUp Dental proprietary data, cited in "Three stats every dentist needs to know about Google reviews," Oral Health Group (2025-09-08)
Learning outcomes
What this study teaches
- Google is still the anchor of dental discovery, but it's no longer the whole story. Consumers are checking an average of six sources before deciding, so a practice watching only its Google listing is watching a shrinking share of the picture.
- Star rating and review count matter, but what's written inside the reviews, especially how staff communicated, reportedly drives satisfaction more than clinical outcomes do. Treat reviews as a service-quality signal to act on, not just a badge to display.
- Patients already trust AI-produced summaries close to as much as they trust the reviews underneath them, and roughly a quarter say they'd decide from the summary alone. Whatever an AI engine says about a practice already carries real weight, even though nobody has independently measured it yet for dentistry specifically.
- Practices are adopting AI fast for scheduling and paperwork and almost not at all for anything a patient sees or hears. That's a defensible clinical choice, but it leaves the patient-facing AI conversation almost entirely unmanaged.
- A basic, unglamorous access problem, getting someone on the phone, still keeps a meaningful share of patients from booking at all. Digital discovery gains mean less if the booking step right behind it is still a bottleneck.
Honest limits
What this does not yet settle
- No large-scale, methodologically transparent primary survey of U.S. dental patients' search and discovery behavior (search engine versus Google Maps versus review sites versus AI answers) currently exists. Widely circulated figures like "71% search online before scheduling" trace back to marketing blogs that disclose no sample or method, so they were left out of this study rather than cited as fact.
- There is no independent census of the AI-answer layer for dental or healthcare queries specifically, meaning no measurement yet of how often ChatGPT, Google's AI Overviews, or Perplexity actually surface or recommend a given dental practice, or on what basis. The AI-trust and AI-usage figures here come from general local-business research (BrightLocal), a general-purpose visibility framework (the KDD 2024 GEO paper), and dental-chatbot answer-accuracy studies that test information quality rather than discovery or ranking behavior. This layer is a genuine frontier, not a measured fact, for dentistry specifically.
- Peer-reviewed research exists on how DSO and private-equity consolidation is reshaping dental market dynamics, but the full text sits behind a paywall and its specific figures could not be verified this pass, so they were left out.
- Official U.S. dental-visit utilization-rate pages could not be reached this pass, so current visit-rate figures could not be independently confirmed.
- Regulatory specifics for dental marketing, including FTC fake-review enforcement in a health context and state-level AI-disclosure requirements for patient-facing dental chatbots, were not covered in this pass and would need a dedicated follow-up.
This is a synthesis of dated, attributed evidence, not a census. The AI-answer layer in particular has no independent, Nielsen-grade measurement yet, so readings of it are directional and named as a frontier, never presented as settled.
Straight answers
Frequently asked questions
Is Google still the main way patients find a dentist?
It's still the leading source, but its share is shrinking fast. BrightLocal's 2026 Local Consumer Review Survey found Google's share as a review or recommendation source fell from 83% to 71% in a single year among general local-business consumers, while use of AI tools like ChatGPT for local recommendations jumped from 6% to 45%, now the third most-used source behind Google and Facebook. The average consumer now checks six different sources before choosing a local business, so a practice watching only its Google listing is watching a shrinking share of a wider picture.
Do patients actually trust what AI tools say about a business?
Yes, and the trust is already substantial. BrightLocal found 40% of consumers trust AI platforms for business recommendations against 32% who don't, 42% trust AI recommendations about as much as traditional reviews, and 82% read the review summaries an AI engine writes when deciding on a business, with 23% saying they'd rely on that summary alone. This is general local-business consumer research, not a dental-specific study, but there is no evidence dental search behaves differently.
How reliable are AI chatbots when patients ask them dental questions?
Inconsistent. A study in the Journal of Prosthetic Dentistry testing five chatbots on geriatric denture questions found Google Gemini scored highest on accuracy and Microsoft Copilot lowest, that Copilot scored highest on empathy while ChatGPT-3.5 scored lowest, and that accuracy and empathy were negatively correlated, meaning no chatbot tested was reliably both accurate and warm. A separate BMC Oral Health study found ChatGPT-4 and Gemini performed comparably to the World Dental Federation's own answers on completeness and clarity, so the picture is mixed rather than uniformly good or bad.
Are dentists using AI with patients yet?
Barely, and mostly not on purpose. The ADA Health Policy Institute's dentist panel survey found 43.3% of U.S. dentists use AI for at least one practice task, but that adoption concentrates on administrative work like imaging, insurance verification, and front-desk check-in, while under 5% use AI for patient treatment recommendations. The same survey found 82.6% of dentists do not plan to use AI for treatment recommendations and 68.3% do not plan to use it to explain clinical findings, citing fear of misdiagnosis and concern about eroding clinical judgment. That caution covers what practices do with AI internally, though, and says nothing about what AI engines are already saying about those same practices when a patient searches.
Has anyone measured what ChatGPT or Google's AI Overviews actually say about a specific dental practice?
No, that gap is real. Every AI-trust and AI-usage figure in the study comes from general local-business research, not dentistry specifically, and no independent census yet exists of how often ChatGPT, AI Overviews, or Perplexity surface or recommend a given dental practice, or on what basis. A peer-reviewed framework presented at KDD 2024 does show that content-level details like statistics, citations, and quotations can raise a source's citation share in generative engine answers by up to 40% in general-purpose testing, but that hasn't been tested on dental content specifically. That measurement gap, not a settled fact, is what the study calls the open frontier for the field.
Provenance
References
- ADA Health Policy Institute, "National Dental Expenditures," analysis of CMS National Health Expenditure data https://www.ada.org/resources/research/health-policy-institute/dental-care-market/national-dental-expenses
- ADA Health Policy Institute, "Dentists Use AI to Make Appointments More Efficient, but Draw the Line at Clinical Decision Making" https://www.ada.org/resources/research/health-policy-institute/dental-practice-research/dentists-ai-usage-and-attitudes
- BrightLocal, Local Consumer Review Survey 2026 (16th annual edition) https://www.brightlocal.com/research/local-consumer-review-survey/
- Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan, Deshpande, "GEO: Generative Engine Optimization," KDD 2024 https://arxiv.org/abs/2311.09735
- Seyed Kian Haji Seyed Javadi, "From Patient Experience to Dental Service Return Visits: A Narrative Review of Strategies for Improving Dental Care Services," Journal of Patient Experience https://pmc.ncbi.nlm.nih.gov/articles/PMC12690051/
- Feizollah A, Byrne M, "Patient perceptions of English dental services before, during and after the COVID-19 pandemic: a sentiment and topic analysis of online reviews," British Dental Journal https://www.nature.com/articles/s41415-025-9099-z
- Sivakumar I, Arunachalam S, Gadde P, Sharan J, "Performance of AI chatbots in responding to geriatric patient questions on denture issues," Journal of Prosthetic Dentistry https://pubmed.ncbi.nlm.nih.gov/41241562/
- Arpaci A, Ozturk AU, Okur I, Sadry S, comparison of ChatGPT-4 and Gemini against FDI oral-health FAQs, BMC Oral Health https://pubmed.ncbi.nlm.nih.gov/40753419/
- Fidancioglu YD, Catak T, "The role of social media in parents' approaches to dental treatment procedures under general anesthesia and sedation," BMC Oral Health https://pubmed.ncbi.nlm.nih.gov/41629922/
- RevUp Dental proprietary data, cited in "Three stats every dentist needs to know about Google reviews," Oral Health Group https://www.oralhealthgroup.com/features/three-stats-every-dentist-needs-to-know-about-google-reviews/
Every measured figure is dated to its capture and tagged with an evidence tier. Every cited work is real and locatable. Where an engine could not be captured this round, it is named as uncaptured, not estimated. Small-sample readings are labelled as directional.
Social Media Is Already Shaping High-Stakes Decisions
A cross-sectional survey of 385 parents in Turkey found 80% had searched social media for information about general anesthesia or sedation before their child's dental treatment, and 64.6% said that information influenced their treatment decision. Social media use was tied to a statistically significant drop in reported anxiety (from 7.18 to 6.09) and a rise in confidence (from 5.31 to 6.82). But 60.1% of parents were also exposed to negative content along the way, rising to 66.7% among heavy users spending three or more hours a day on the platforms.
This is one study, in one country, about one specific clinical decision. It's also a real, measured data point showing that even for a high-stakes choice, most parents went looking on social platforms first, and it changed their minds, for better and for worse depending on what they happened to find. The discovery layer for dental decisions clearly runs well past search engines.