Do appointment reminders actually reduce no-shows?

Most claims about no-shows come from software vendors quoting each other. This one does not have to: it has been through a Cochrane systematic review, which is about as good as evidence gets outside a lab.

What the evidence says

Roughly eleven percentage points, from a text message. The review pooled several randomised controlled trials with thousands of participants, which means the finding is not an artefact of one clinic or one country.

The caveats, stated properly

This research is from healthcare settings, and the messages studied were SMS. It is reasonable to expect the mechanism to carry across to a salon, a garage or a consultancy — people forget appointments everywhere — but it has not been measured there, and we are not going to pretend otherwise.

It is also an average. If your no-show problem is mostly people who never intended to come, a reminder will not fix it. Reminders work on forgetting, not on ambivalence.

What to do with it

The practical design that follows from the evidence is dull and effective. Send the reminder on the channel the person actually reads — in this region, that is usually WhatsApp rather than email. Ask a question they can answer in one tap, because "still good for tomorrow at 2?" gives you a cancellation you can refill, whereas "this is a reminder" gives you nothing.

And make the cancellation useful. A slot that frees up 18 hours ahead is worth real money if someone offers it to a waiting customer within the minute, and worth nothing if it is noticed the next morning.

Where deposits fit

Deposits are the other lever, and they work through a different mechanism: they make not turning up cost something. They also cost you bookings from people who would have come but resent being asked. The reasonable middle is to ask for one selectively — first-time customers, long lead times, high-value slots — rather than universally.