Cancellation rate is the most quoted and least standardised measure in a service business. Two studios can both describe a large share of their appointments cancelling and mean entirely different things, because one counts only same-day call-offs and the other counts every change of plan, including a move booked three weeks out.
The number is not wrong in either case. It is answering a different question, and the label hides which one.
Three events that get called the same thing
A client who calls the day before and moves to next month has cancelled a slot. A client who does not arrive has also left a slot empty, but gave no warning. A client who rebooks during the appointment has not cancelled anything at all, though the record may look like a new booking replacing an old one.
Each event has a different cost. A warning gives you a chance to refill. Silence does not. A rebooking inside the visit is the outcome you want, and counting it as churn punishes success.
The first decision is which events the rate will include, written down before the month starts.
Choosing the denominator
Most of the variation between reported figures comes from the bottom of the fraction, not the top.
Divide cancellations by appointments booked, and a busy month with a long booking horizon looks better than a quiet one, because the horizon inflates the count of bookings without changing the number of cancellations.
Divide by appointments scheduled to occur in the period, and the figure describes the calendar you actually had to fill. That is usually the more useful version for a studio, because it matches the hours at risk.
Divide by hours rather than appointments, and long services carry more weight. That matters in a business where a single long appointment can strand an afternoon.
Pick one, state it, and keep it.
Lead time is the split that changes decisions
A single cancellation figure tells you that slots were lost. The lead time tells you what to do about it.
Group cancellations by how far ahead of the appointment they were made: same day, the day before, earlier in the week, and well ahead. Each group points at a different fix. Same-day losses respond to a confirmation sent earlier and a waitlist that can be offered quickly. Losses made a week out respond to a reminder that arrives before the client's week fills up, and to a rebooking conversation at the end of the previous visit.
Same-day and no-show losses in the same bucket point somewhere else again: a client who never intended to come, or one who could not reach you to say so.
Why the figure moves when nothing changed
Several things shift the rate without any change in client behaviour. A month with more long weekends has more cancellations, because plans change. A month with a marketing push has more first-time bookings, and first-time clients cancel differently than regulars.
This is why the figure should be read against the same month a year earlier, and why a single month is a poor basis for a policy change. Two consecutive periods moving the same direction is a signal; one is a data point.
There is also a reporting trap. When reminders begin going out by text, more clients call to move an appointment rather than simply not arriving. The cancellation count rises and the no-show count falls. The studio has improved, and the headline has worsened. Read the two together or the improvement looks like damage.
Reading the figure next to what caused it
Elevoi answers calls at any hour and books against live availability, and every booking produces a text confirmation. That confirmation is often the moment a client realises a conflict, which turns a silent no-show into an early cancellation you can refill. The call log keeps the reason a slot moved, because the client tells the assistant, and the record stays attached to the booking.
What Elevoi does not do is publish a target rate. There is no benchmark value to hit, and your own month-over-month figures are the comparison.
A short routine
Once a month, pull the appointments that did not happen. Tag each one: cancelled with notice, cancelled late, no-show, moved. Count them against the hours scheduled in the period, not against bookings made. Then look at where they cluster: which weekday, which service, which provider's book.
Pick one cluster and one change. Then measure again with the same tags. Most studios find the same-day group is the one worth attacking first, because it is the group a reminder and a waitlist can actually reach.
Start from the salon industry statistics hub, or read the staff utilization benchmarks statistics explainer for the other half of the capacity picture.