Open ChatGPT in a new tab. Type: "best family dentist in Calgary." Now look at the names that come back.
If you're a dentist in this city, there's a roughly 90% chance your practice isn't one of them. That's not because your work isn't good. It's because ChatGPT, Claude, Perplexity, and Gemini don't read websites the way Google does, and most dental sites are built for Google.
This matters more every quarter. A growing share of new-patient research now starts with an AI tool, not a search bar. By the time prospective patients reach your contact form, they've already narrowed their list to two or three names, and those names came from somewhere.
What AI tools actually look at
When ChatGPT recommends a dentist, it's not running a fresh Google search. It's pulling from a mix of three things: structured data on your website (schema markup), citations of your practice across third-party sources (review platforms, news sites, directories), and entity associations it's built up about you over time.
Practices that get cited tend to do three things well, and most practices do none of them.
Fix one: actually publish Dentist or MedicalBusiness schema
Schema markup is structured data that tells AI tools exactly what your business is, where it operates, what services it offers, and who runs it. Most dental sites have either no schema at all or a generic LocalBusiness schema that an SEO plugin auto-generated and never got refined.
The version you want is Dentist (a subtype of MedicalBusiness). At minimum it should declare your practice name, address, phone, hours, accepted insurance, languages spoken, and the specific services you offer (cleanings, implants, Invisalign, pediatric, sedation, whatever applies). If you have multiple dentists, each one gets a Person schema with their credentials, education, and specialties.
You can validate your schema using Google's Rich Results Test or Schema.org's validator. If your dev or SEO person tells you "we already have schema," ask them to send you the JSON-LD output and run it through the validator yourself. About half the time, what's on the page is broken or outdated.
Fix two: review velocity matters more than total count
Most dentists obsess over their total Google review count. AI tools care more about velocity, how recent and how steady the reviews are. A practice with 80 reviews and three new ones every month outranks a practice with 400 reviews where the last new one was eight months ago.
The mechanical fix: build a system that asks every patient for a review the day after their appointment. The cleanest version is a simple SMS sent automatically through your practice management software (Dentrix, Open Dental, Curve, etc.) with a direct link to your Google Business Profile review page. Don't ask in the office. Don't hand them a card. Send a text the next morning when they're sitting on their couch and the visit was pleasant.
Aim for 8-12 new reviews per month. That's enough velocity to signal to AI tools that your practice is active and that recent patients are happy. It's also enough to push you past the 4.7-star threshold where AI tools start preferring practices in their recommendations.
Fix three: get cited by sources AI tools already trust
The single fastest way to start showing up in ChatGPT recommendations is to be mentioned by name on third-party sites that AI tools have already crawled and trusted. For dental practices in Canadian markets, the highest-leverage sources are: your provincial dental association directory, RateMDs, your local chamber of commerce, news mentions in local press (CBC, Global News, your city's daily paper), and dentist-specific aggregators like 1-800-Dentist or Opencare.
Your name needs to appear consistently, same practice name, same address format, same phone number, across all of these. Inconsistency kills you. "Smith Family Dentistry," "Dr. Smith Dentistry," and "Smith Dental" being three different listings means AI tools see three different entities and trusts none of them.
Spend an afternoon auditing every directory listing you can find for your practice. Pick one canonical name and address format. Update everything to match. This sounds boring because it is. It also works.
What this gets you
Three to six months after implementing the above, practices we've worked with start showing up in ChatGPT and Perplexity responses for queries like "best dentist near me" and "[city] family dentist", usually for the first time. The dentists who got there earliest are the ones who fixed schema, built review velocity, and cleaned up their citations before their competitors did.
None of this is glamorous. None of it requires you to spend more on Google Ads. It requires about ten hours of mechanical setup and then a system that runs itself. Most practices won't bother. That's the opportunity.
One thing not to do
Don't let an SEO agency convince you to publish 40 thin "Why You Need a Dental Cleaning" blog posts. AI tools have gotten very good at recognizing low-quality, mass-produced content and they actively deprioritize practices associated with it. One well-researched neighborhood guide ("Dental services in [your neighborhood]: what to know about insurance, wait times, and after-hours emergencies") beats forty generic blog posts every time.
Quality over volume. Specificity over breadth. The practices that win the AI search era are the ones that look like real, specific, locally-rooted businesses, because that's what AI tools are trying to recommend.