Patient questions
The questions an implant patient asks before calling anyone
Patients do not search for keywords. They ask sentences, and the sentences repeat across cities. Build the library once.

The owner’s takeaway
Across markets, implant patients ask a recognisable set of questions in a recognisable order. The library is finite, which makes it testable.
An implant patient moves through stages: what is this, am I a candidate, what does it cost, who does it well near me, and what happens if it fails. Each stage produces a sentence, and each sentence is a place where an answer either names a practice or does not.
The set is smaller than most practices expect. Twenty to forty questions cover the decision in one metro.
A finite, testable list
- Collect
Front desk list
Ten verbatim pre-booking questions.
- Add
Clinical questions
What the team answers in consults.
- Test
Two engines
Record names and sources per question.
- Freeze
90 days
Same list, monthly recheck.
Twenty to forty questions cover one metro's implant decision.
Five stages, five question shapes
- Understanding: what is the difference between an implant and a bridge?
- Candidacy: is bone grafting always needed before an implant?
- Cost: what does a full arch cost in [city] with insurance?
- Selection: who is the best full-arch dentist near [city]?
- Risk: what happens if an implant fails in the first year?
The first three are content problems. The fourth is where a practice is named or not. The fifth decides whether a patient trusts the answer enough to call.
How to build the library in an afternoon
Start with the front desk. Ask for the ten questions patients ask before booking, verbatim. Add the questions the clinical team answers in consults. Then run those sentences through two engines and record which ones return a named practice.
Keep the list stable for 90 days. Changing questions mid-quarter resets the baseline and makes every trend unreadable.
Using the library without turning it into a content farm
One page per stage, not per keyword. A candidacy page that answers the twelve candidacy questions beats twelve thin pages. The model cites the page that resolves the question, not the page that repeats it.
Your next move
Start with five of them
The snapshot tests the highest-intent questions first and returns the wording, not just a score.
Free AI snapshotRun the five-question snapshot

