AI Assistants Are Recommending Dentists to Patients, and Most Practices Cannot See It, New Study Finds

A 90-day study of a boutique dental practice found new patients arriving through ChatGPT while website analytics showed nothing, alongside AI assistants repeating false facts and routing pricing questions to a competitor.

Seattle, Washington - A new case study from AI visibility firm Halogen has documented something most dental practices have no way to detect: AI assistants like ChatGPT, Google Gemini, and Claude are actively recommending dental practices to prospective patients, and those referrals are largely invisible to the analytics tools practices rely on.

Over a 90-day engagement with a boutique dental practice in Hawaii, five new patients reported finding the practice through ChatGPT. None of them appeared in the practice's website analytics as anything other than ordinary direct traffic. The only reason the practice knew was a single question added to its intake form: "How did you hear about us?" with an AI option included.

"For twenty years, being findable meant being on Google," the practice's owner said. "Nobody told me the rules had changed. I had no idea a patient could ask a chatbot for a dentist and get sent straight to us, or straight past us."

The study, published by Halogen, tracked fifteen common patient questions across ChatGPT, Google Gemini, and Claude over three months. Its findings challenge how local businesses think about being discovered online.

Patients Arrive, But Traffic Does Not Move

The five new patients represented roughly $5,000 in patient value over three months, using a conservative estimate. Industry figures place the lifetime value of a general-practice dental patient between $6,000 and $10,000, and higher for implant and cosmetic cases.

Yet the practice's website traffic never rose. Nineteen months of analytics showed daily sessions holding roughly flat, with no summer surge to explain the new arrivals. The reason, the study concludes, is structural: when an AI assistant recommends a practice, it often hands the patient a name and phone number directly. The patient calls without ever visiting the website, so the referral leaves no trace in standard analytics.

What the Assistants Say Changes From Day to Day

The study found that AI visibility is not stable. One question naming the practice changed its answer on ChatGPT ten times over five weeks, alternating between recommending the practice and omitting it entirely, all while the practice's website was never touched. On one day, ChatGPT returned no mention of the practice at all for a question that named it directly.

Because the answers shift constantly, the study argues that a one-time check tells a practice almost nothing, and that meaningful measurement requires ongoing monitoring rather than a single audit.

The Assistants Get Facts Wrong

Reading the actual answers, rather than only tracking whether the practice appeared, surfaced errors the practice would never have caught on its own. Google Gemini described a dentist as the practice's pediatric provider; that dentist had left the practice, and the assistant was drawing on outdated reviews. ChatGPT told patients the practice accepted a financing option it does not offer, pulled from a third-party directory. Neither claim originated with the practice.

In one of the study's most striking findings, when asked what dental implants cost at the practice, ChatGPT noted that no pricing was available on the practice's own site and then filled the gap by citing a competing dental practice in the same market. On one of the highest-intent questions a prospective patient can ask, the assistant was pointing a ready-to-book patient toward a competitor.

A Market Largely Unaware

Three competing practices in the same market registered at or near zero visibility across all three assistants for the entire study period. Nothing external predicted which practices appeared and which did not, not size, not location, not website quality. The study notes that two of those competitors are effectively invisible to AI in their own market and almost certainly do not know it.

"Seeing the actual answers was the part that got me," the practice owner said. "A chatbot was telling patients we take a kind of financing we don't, and naming a dentist who left. I'd never have caught it, because I never thought to ask what these things were saying about us."

Why It Matters Now

The study arrives as consumer behavior shifts toward asking AI assistants for recommendations the way people once relied on search engines. For local service businesses, and dental practices in particular, the answer an assistant gives can represent a patient's entire consideration set, delivered as a single response with no second page of results.

The full case study, including methodology and detailed findings, is available at www.halo-gen.ai. At the participating practice's request, the practice and its dentists are not named.

Co-Founded by Amazon Veterans Who Built at Scale

Mario V Adoc brings 18 years of Amazon experience to Halogen. He launched Twitch Prime, delivering hundreds of free games and exclusive in-game content to millions of subscribers worldwide, and built the operational platform powering Prime Video's live sports portfolio, including the NFL, NBA, English Premier League, UEFA Champions League, Roland-Garros, and the US Open. He co-founded Halogen with CTO and fellow Amazon alum Taylor Manners to apply that same operational rigor to the emerging discipline of AI brand visibility.

Start a Free AI Visibility Check

Brands and agencies can assess their current AI search visibility at no cost using the Halogen Presence AI Visibility Check at app.halo-gen.ai/ai-visibility-check. The check identifies how a brand currently appears across major AI platforms and surfaces a prioritized list of recommendations.

Paid plans, including an entry-level $99 Audit with credit toward a subscription, are available at halo-gen.ai.

Halogen
1545 NE 90th St
Seattle, WA 98115

Halogen is an AI search visibility company headquartered in Seattle, Washington. Its platform, Halogen Presence, tracks how brands appear inside the responses of generative AI assistants, identifies the citation sources those answers are built from, and provides the structured recommendations and content tools to improve that visibility. Halogen was co-founded by Mario V Adoc and Taylor Manners. Learn more at halo-gen.ai.

This release was published on openPR.