Overview metrics
The appointment funnel, completion, no-shows, and how many bookings the assistant produced.
Five numbers, all over a period you choose. None of them is a vanity metric; each answers a question you would otherwise guess at.
The appointment funnel
Booked → confirmed → completed, with what fell out at each step. The shape tells you where you are losing people, which is almost never where clinics assume.
Completion rate
Of appointments that reached a conclusion, how many were completed. This is your headline operational number.
No-show rate
Of resolved appointments, how many the patient simply did not attend.
This number is only true if your team marks no-shows as no-shows rather than cancelling them to be kind. If it looks implausibly low, that is the first thing to check.
When it is genuinely high, there are three levers and they are worth trying in this order: self-service links, reminder count, and how many of your patients have an email address at all.
Bot success rate
Of appointments the assistant produced, how many were completed. It answers the question everyone actually has about an AI assistant: are these real patients?
Compare it against your overall completion rate rather than against 100%. If bot-sourced appointments complete at roughly the clinic average, the assistant is bringing you ordinary patients — which is the good outcome.
Modality and distribution
Online against in-person, and how appointments spread across your specialties. The specialty split is the one to read before you hire — a specialty with a long waitlist and a small share of appointments is a bottleneck, not a niche.
Common questions
- Why is the funnel narrower than my appointment count?
- The funnel counts appointments that reached a conclusion in the period. Visits still in the future are not yet anything.
- Can I compare two periods?
- Change the period and read it again. What matters more than any single figure is the direction over three of them.