We measured 100 Canadian dental keywords covering 242,500 monthly searches. 55% return an AI Overview, but the distribution is uneven: 94% of cost and insurance questions, 92% of comparison questions, and only 10% of searches naming a city. AI answers took the research stage; the map pack still owns local intent.
Key takeaways
The short version, for anyone who does not read to the end.
- 55% of the Canadian dental keywords we sampled return an AI Overview, accounting for 61% of the search volume.
- Cost and insurance questions return one 94% of the time. Comparison questions 92%.
- Searches naming a city return one only 10% of the time. The map pack still owns local intent.
- The two rarely co-occur: of keywords showing a map pack, only 17% also showed an AI Overview.
- 29 of the AI Overview keywords sit at keyword difficulty 0, together worth 46,200 monthly searches.
- Check whether your site blocks AI crawlers before anything else. Cloudflare and several hosts block them by default.
What we measured
In October 2026 we pulled the search engine results features for 100 Canadian dental keywords with at least 500 monthly searches each — 242,500 monthly searches in total, spanning general dentist searches, implants, aligners, root canals, whitening and emergency care.
For each keyword we recorded whether Google returns an AI Overview, whether it returns a local map pack, the search volume and the keyword difficulty. The question was simple: has the generated answer actually arrived in dental search, and if so, where.
It has, and the answer to "where" is more specific than the industry discussion suggests.
Which dental searches return an AI Overview?
Overall, 55 of the 100 keywords returned one — 61% of the sampled search volume. But splitting by what the patient was actually asking produces a much sharper picture.
Cost and insurance questions return an AI Overview 94% of the time: 15 of the 16 we sampled. "Invisalign cost", "root canal cost", "dental implant cost", "is Invisalign covered by insurance in Canada" — all answered above the results.
Comparison and explanation questions return one 92% of the time. "Invisalign vs braces", "what is a root canal", "is Invisalign faster than braces", "benefits of Invisalign over braces".
Searches naming a city return one 10% of the time. Of 30 city keywords — "dentist Toronto", "dentist Mississauga", "dental implants Calgary", "teeth whitening Edmonton" — only three returned an AI Overview.
"Near me" searches fall in between at 36%, and the split within them is informative: "emergency dentist near me" and "dental implants near me" return one, while "dentist near me", "invisalign near me" and "teeth whitening near me" do not.
Do AI Overviews replace the map pack?
No, and the data is unusually clean on this point.
Of the 48 keywords in our sample that return a local map pack, only 8 also return an AI Overview — 17%. Of the 55 keywords that return an AI Overview, 47 have no map pack at all.
They are answering different questions. When Google reads local intent, it shows businesses. When it reads research intent, it generates an answer. A practice is not competing in one arena that has changed shape; it is competing in two, and most dental marketing advice only covers one of them.
This is also why the panic framing is wrong. "Dentist near me" at 35,000 monthly Canadian searches returns a map pack and no AI Overview. The single highest-volume dental search in the country is unaffected, and a complete Google Business Profile still decides it.
What this means for which pages to build
The pages that now sit under an AI Overview are precisely the pages most clinic websites do not have.
In our earlier study of 213 Canadian dental clinics, the median practice covered 55 keywords and had four or five pages in total — a homepage, a combined services page, an about page and a contact form. None of those answer a cost question or a comparison question.
So the opportunity is unchanged in substance and more urgent in timing. Build the cost page, the comparison page and the candidacy page. They were already the highest-potential pages by search volume. Now they are also the ones determining whether a practice is cited when Google answers the question directly.
29 of the AI Overview keywords in our sample sit at a keyword difficulty of 0, together worth 46,200 monthly searches. The barrier is not competition. Our AI Search Optimization for Dentists page sets out how that work runs.
How does a clinic become a cited source?
Nobody outside Google can describe the selection process precisely, and anyone claiming otherwise is guessing. What can be said is what makes a page usable as a source, and it is mostly unglamorous.
Answer the question directly, near the top, in a short self-contained passage. A generated answer is assembled from extractable statements, not from a page that builds to a conclusion after six paragraphs of preamble.
Use the question as the heading, phrased the way a patient phrases it. "How much does Invisalign cost in Canada?" is findable in a way that "Understanding Your Investment" is not.
Give specifics — ranges, conditions, what changes the answer. Vague content is not citable content, and the clinics that publish real numbers are the ones with something to quote.
Describe the practice in structured data so it resolves as a known entity rather than being inferred, and keep the name, address and phone consistent everywhere they appear.
Is your site even readable by AI crawlers?
Check this before any content work. It is the most common blocker and it has nothing to do with writing.
A site disallowing GPTBot, ClaudeBot or Google-Extended cannot be cited by the systems those crawlers feed, regardless of how good its pages are. We later audited 195 Canadian dental websites and found only 2.1% doing so, so this is a rare failure rather than a widespread one — worth five minutes to rule out, not worth paying for as a service.
This site was in exactly that position earlier this year — the host's managed robots.txt blocked eight AI crawlers, and nobody had chosen that. Opening retrieval access while keeping training access closed is a reasonable middle position, and it is a setting rather than a project.
Read your live robots.txt rather than the file in your repository. Managed platforms frequently override it.
How should this be measured?
Not with screenshots. AI Overviews are generated per query and vary between users and days, so a screenshot proves only that something happened once.
Track citation counts across platforms over time — Ahrefs and similar tools now index this. Watch branded search volume, which tends to move when a practice starts being named in generated answers. And keep measuring tracked calls and booked appointments, because a citation that produces no patients is not a result.
Expect the numbers to be small at first and to be unevenly distributed across platforms. A site can be cited repeatedly by one assistant and not at all by another, and the reasons are not externally visible. Call Tracking for Dentists is what connects any of it to a booked appointment.
What we would not do
Rewrite existing pages to sound quotable at the expense of being accurate. Health content that overstates certainty to seem authoritative is a professional risk before it is a marketing one, and that risk sits with the clinician rather than the agency.
Abandon local SEO. The highest-volume dental searches in Canada still return a map pack and no AI Overview, and the map pack still sits above everything.
Pay for guaranteed AI Overview placement. It is not a placement and it cannot be guaranteed.
AI Overview coverage by search intent — 100 Canadian dental keywords, 242,500 monthly searches (Ahrefs, October 2026)
| Search intent | Keywords | With AI Overview | Share | Monthly searches |
|---|---|---|---|---|
| Cost and insurance | 16 | 15 | 94% | 30,000 |
| Comparison and questions | 13 | 12 | 92% | 19,700 |
| “Near me” | 14 | 5 | 36% | 57,600 |
| City named | 30 | 3 | 10% | 27,600 |
| Other | 27 | 20 | 74% | 107,600 |
| All sampled | 100 | 55 | 55% | 242,500 |
FAQ
On most informational ones. In our 100-keyword sample, 55% returned an AI Overview, rising to 94% for cost and insurance questions and 92% for comparison questions. Searches naming a city returned one only 10% of the time.
From research-stage pages, likely some. From local searches, much less — the highest-volume dental search in Canada, "dentist near me" at 35,000 monthly searches, returns a map pack and no AI Overview. The impact is concentrated on cost and comparison content.
There is no published mechanism, and anyone guaranteeing placement is describing something they do not control. What helps is answering the question directly near the top of the page, using the question as the heading, giving specifics rather than generalities, and making sure AI crawlers are not blocked from the site.
Check the live robots.txt rather than the file in your repository — managed hosts frequently override it. Cloudflare and several platforms block GPTBot, ClaudeBot and Google-Extended by default, which prevents citation by the systems they feed.
No. Of the keywords in our sample returning a map pack, only 17% also returned an AI Overview. Local intent still resolves to businesses. The two compete in different places.
By citation counts tracked over time across platforms, alongside branded search volume and the usual tracked calls and booked appointments. Not by screenshots — AI Overviews vary by user and by day, so a single capture proves nothing durable.