Automated reminders are one of the few clinic changes that pay back quickly, and one of the most commonly implemented badly. A clinic switches them on, missed appointments fall somewhat, and everyone concludes the job is done. The gap between somewhat and substantially is mostly made of small decisions nobody revisited.
This is about those decisions: what a reminder should say, when it should arrive, what to do with the replies, and why the underlying data matters more than the message.
A reminder is a data problem first
Before any question about wording, there is a question about whether the message can arrive at all.
A reminder goes to a contact detail captured at some earlier point, usually by whoever was on the desk, often while being interrupted. If that number is wrong, nothing announces it. The message does not bounce in a way anyone sees. It simply goes nowhere, and the clinic learns about the problem when a chair sits empty.
This has a practical consequence. The highest return improvement to a reminder system is usually not better copy, it is making contact details easy to correct at every visit rather than only at registration. A patient who changed number two years ago will not volunteer that; they will answer if asked while paying.
Clinics that want to know the size of this problem can look for it directly. Count how many active patient records have a number that has never been updated since registration, and how many have no mobile number at all. In most practices the answer is larger than expected and explains part of the no show rate on its own.
Timing, and why one reminder is rarely right
A single reminder has to do two incompatible jobs: give enough notice to rearrange, and arrive close enough to be remembered.
| When | What it achieves | What it misses |
|---|---|---|
| At booking | Confirms details are right while the patient is present | Nothing, but it is not a reminder |
| One week before | Time to rearrange, so the slot can be refilled | Forgotten by the day itself |
| One to two days before | Close enough to act on | Too late to fill a cancelled slot easily |
| Morning of | Catches the genuinely forgetful | No chance of refilling at all |
The week before message is the one clinics skip and the one that protects the calendar, because it is the only point at which a cancellation is still commercially useful. A patient who cancels seven days out leaves a slot you can fill. The same cancellation at eight in the morning leaves a gap.
For most practices, two messages work better than one or three: a week ahead, and a day ahead. More than that starts to read as nagging, and patients who feel nagged stop reading the messages entirely, which removes the benefit for the appointments that actually needed it.
Longer appointments deserve different treatment. A two hour surgical slot is worth confirming by a person rather than a message, because the cost of it going empty is much higher and the conversation may also surface a concern the patient has not mentioned.
What the message should actually say
Reminder copy tends to be either too sparse to be useful or padded with things nobody reads.
The useful content is short. Who it is from, which is the clinic name rather than an unexplained number. When, with the day of the week as well as the date, because people think in days. Where, if the practice has more than one location. And what to do if the time does not work, with a route that is actually easy.
What to leave out matters as much. Do not include the treatment being performed. A message describing a procedure may be read by someone other than the patient, and the clinic has no way to know who has access to that phone. Confirming the appointment does not require disclosing its purpose.
The reply route is where most reminders fail. A message that says to call during opening hours has asked the patient to do the hardest possible thing. If replies are accepted, somebody has to read them, which is a staffing decision rather than a software one. A reminder system nobody monitors is a broadcast, and patients work that out within two cycles.
Consent, preference and the awkward middle
Patients differ in how they want to be contacted, and treating that as a single switch is a mistake.
Ask at registration and record the answer. Some patients want a message, some want none, some want a call for anything longer than a check up. The distinction that matters is between appointment reminders, which most people welcome, and anything promotional, which is a separate permission and should be recorded separately.
Respect the preference everywhere. A patient who opted out of messages and receives one anyway has learned that the clinic's records do not govern its behaviour, which is a worse impression than never having asked.
Check what happens when an appointment is moved after the first reminder has gone out. Systems that send a second message without referencing the first produce confused patients, and the resulting phone calls cost more desk time than the reminder saved. This is worth testing during a demo rather than discovering live.
Where reminders run from the same calendar the clinic books in, a rescheduled appointment updates the reminder rather than adding to it, which removes the problem rather than managing it.
What to do with the ones who still do not come
Reminders reduce missed appointments; they do not eliminate them, and the residue is worth understanding rather than absorbing.
Record the miss on the patient record rather than deleting the appointment. A deleted no show is information destroyed: you lose the ability to see a pattern, and the next person booking that patient has no context. This is not about penalising anyone, it is about knowing whether a patient has missed once or four times, which calls for different responses.
Look at what the misses have in common before concluding anything about patients. First appointments are missed more than continuing ones. Mondays and the first slot after lunch are missed more than mid morning. Appointments booked more than two months ahead are missed more than recent ones. If your pattern matches those, the response is a scheduling adjustment rather than a patient conversation.
Distinguish a miss from a late cancellation in the record, even when the operational effect is similar. A patient who called an hour before behaved differently from one who did not call at all, and treating both the same removes the clinic's ability to tell a disorganised patient from an inconsiderate one. The first responds to a better reminder; the second may not.
For the small group who repeatedly miss, a short phone call asking whether the times offered work for them is more productive than a policy. Occasionally it surfaces something practical, such as a patient who cannot take time off and has been booking slots they knew were difficult.
Recalls are a different job
Clinics often run recalls through the same machinery as reminders, and the two are not the same thing.
A reminder concerns an appointment that exists. A recall concerns one that does not: the patient is due, nothing is booked, and the message has to create an action rather than confirm one. That difference changes everything about how it should be written and measured.
Recall messages fail differently too. A reminder that does not arrive costs one slot. A recall that does not arrive costs a patient who quietly stops attending, and nobody notices because nothing was in the diary to go empty. This is the failure that shrinks a practice slowly, and it is invisible in any report that counts appointments.
The useful measure for recalls is therefore the opposite of the reminder measure. Rather than counting messages sent, count patients due who did not book, month by month. That list is actionable, it is usually short, and the patients on it already chose your clinic once.
Recalls also tolerate a slower, more personal approach. A message, then a second one a few weeks later, then a phone call for patients the clinic particularly wants to keep, is proportionate. The same escalation applied to an ordinary appointment reminder would be excessive.
Measuring whether it worked
Clinics switch reminders on and rarely check the effect properly, which makes it hard to justify improving them.
Take a simple baseline before changing anything: missed appointments as a percentage of booked appointments, over at least two months, because one month is noise. Then compare the same figure a quarter later.
Expect a real but unspectacular improvement, and be suspicious of a dramatic one, which usually means the baseline was measured differently rather than that something changed. The second thing to watch is cancellations with notice, which should rise as misses fall. That is the system working: a cancelled slot you can refill is a much better outcome than an empty chair.
If you want to see how booking, contact details and reminders sit in one place rather than three, the scheduling overview covers it, and you can book a demo to test the reschedule case above before committing.
