Analysis for clinics, practices and anyone booking appointments.

Did not attend, cancelled late and we filled the slot, and we moved it ourselves are all written into the same column. Only the first cost you a session.

Separated, a headline of 22.4% becomes 11.4% in one worked example. That is not a cosmetic improvement: it changes whether the problem is worth an intervention at all.

How far ahead an appointment was made is typically the strongest predictor in the file, and it is the field most exports bury or omit.

Where it exists, the pattern is usually monotonic across bands and holds within every clinician and appointment type, which is what distinguishes a real driver from a coincidence. If your export lacks the booking timestamp, it is worth going back to the system for it.

Dividing attended sessions by booked sessions produces a figure that cannot fall below a hundred per cent except through no-shows, and therefore says nothing about capacity you failed to sell.

The denominator has to come from the rota, with blocked administrative time excluded and walk-ins kept out because they occupy a slot that was never offered.

One row per appointment with an outcome status, and the field that matters most: the date the appointment was booked, not only the date it was for. Most booking systems can add it and most default exports leave it out.

A rota or session template turns booked volumes into utilisation. Without it, attendance and demand patterns still work and utilisation is reported as unavailable rather than divided by whatever number happens to be present.