Demand was mapped by patient scenarios, not by a loose keyword list
The work started not from a generic keyword dump, but from real intents: when a person is already searching for an appointment, when they are still comparing options, and when they enter through a symptom or a question.
That clustering made it clear where a dedicated service page was needed and where an article, FAQ, or supporting page would do a better job. It removed a common local SEO mistake: pushing too many different formulations onto one weak page.