What actually reduces no-shows, according to the trials

Where the numbers come from

Almost all usable research on no-shows comes from healthcare. That makes sense: hospitals and clinics have thousands of appointments a month, records that capture everything, and a direct financial incentive to fix the problem. A salon with thirty appointments a week does not have that dataset.

Which means you cannot transplant the percentages below onto your own practice. What does transfer is the direction: which measures made a difference under controlled conditions, and which did not. That is more than most advice on this subject can claim.

What a reminder does, and how much the channel matters

Start with the basic question: does reminding help? Yes, and that is the best-supported conclusion in this entire article. The more interesting question is how you remind.

Geneva settled that one properly. Researchers split 6,450 patients at a university primary care clinic into two groups: one got a text message, the other got a phone call. The result: 11.7 per cent no-shows for SMS, 10.2 per cent for phone calls. That difference was not statistically significant, and it met the pre-registered threshold for calling SMS “not worse”.

That is a useful finding, because calling costs many times more. Patients also rated the text message higher. The lesson is not that SMS is magic, but that the expensive channel buys you nothing extra.

The study everyone points at

Now the study you will meet in almost every blog post on this subject, usually without a citation and often summarised wrongly.

In 2015 Hallsworth and colleagues published two randomised trials in British hospitals. In the first, 10,111 patients received one of four text messages five days before their appointment. The outcome:

• the existing message: 11.1 per cent missed appointments;
• a message stating the specific cost of a missed appointment: 8.4 per cent;
• a message making it easier to cancel: no significant effect;
• a message invoking a social norm (“most people do attend”): no effect.

The second trial, with 9,848 patients, confirmed the direction and added something surprising. Stating the specific amount still performed best at 8.2 per cent. Talking vaguely about costs did worse (9.9 per cent). And the variant phrased sympathetically did worst of all: 10.7 per cent.

That is the counter-intuitive finding. The kindest text performed least well. The concrete number performed best, and cost nothing.

And then it did not replicate

If the article stopped here you would have a satisfying takeaway and an incomplete picture. Because in 2023 something happened that you rarely see in the summaries.

Teo and colleagues ran a far larger study at an American veterans’ health system: nearly 50,000 appointments in primary care and another 39,000 in mental health. They tested four variants of exactly this kind of nudge in reminder letters: consequences for yourself, consequences for others, a social norm, and all of them combined.

Not one variant beat the control group. In primary care, 11.1 per cent missed against 11.9 per cent in the control; in mental health, 20.1 against 18.0 per cent. The authors’ conclusion was blunt: reminder letters incorporating brief behavioural nudges were ineffective.

There are differences that might explain it. Letter versus text message. An American veterans’ population versus British outpatient clinics. Costs to the health system versus costs that land on the patient. But the point stands: the reminder itself is the reliable measure, the clever wording is not. Anyone promising to double your attendance on the strength of one sentence in a text message is selling you a study whose follow-up they have not read.

What to do with this in practice

Translated to a practice of ordinary size, in order of confidence:

1. Make sure a reminder goes out at all, and that it arrives. By far the largest effect and the dullest advice. A reminder in the spam folder is not a reminder.
2. Pick the cheapest channel that gets read. More expensive has not been shown to be better. Phoning buys no measurable gain over a message.
3. Make cancelling easy. In the 2015 trial this did not reduce no-shows, but it does turn a no-show into a freed-up slot you can still fill. That is a different gain than the statistic measures.
4. Be concrete rather than sympathetic. Weak evidence, but it costs nothing and the downside is small. “Your appointment is Tuesday at 14:00, cancel via this link” beats a paragraph about how inconvenient this is for everyone.
5. Only consider money once the above is in place. A deposit works, but it also turns away customers who would have come. That is a decision about what kind of business you want to be, not only about a percentage.

Where Calio stands

Calio sends two reminders by default: 24 hours and 1 hour beforehand. Those timings are a practical choice, not a result from the research above; the literature on optimal timing is thinner than the literature on whether to remind at all.

You can change the wording of those messages with workflows, so if you want the concrete time and a direct cancellation link up front, you can have that. Every guest gets their own link to reschedule or cancel without emailing you. That is point three from the list above.

For those who want to go further there is a per-event-type no-show fee. Deliberately in the shape we called reasonable above: the amount is visible before booking, and it is only sent afterwards as a payment request, and only if nobody turned up. Nothing is ever charged automatically.

And because you cannot know whether something works without measuring: your statistics page shows your own no-show rate alongside your cancellations and your conversion. That is the one number in this article that is about your practice.

In short

• Reminding works. Best-supported conclusion here, and also the least exciting.
• The channel matters little. In a randomised trial SMS was no worse than phoning, and far cheaper.
• The clever wording is unreliable. Stating costs helped in 2015 (11.1 to 8.4 per cent), but comparable nudges did nothing at all in a much larger 2023 study.
• Sympathetic phrasing did worse than being concrete. That is the one piece of advice here that runs against intuition.
• Nearly all of this research comes from healthcare. Take the direction, not the percentages.