A cancellation feels like a loss. Someone booked, something came up, the slot is empty and the afternoon has a hole in it.
But an empty slot and a lost customer are not the same thing, and the data says they come apart more often than you’d expect. When someone can’t make the time they booked, cancelling is only about half the story.
So we looked at a year of cancellations and reschedules to see what actually happens when a meeting falls over.
About one booking in eight is cancelled
Across all bookings, 12.2% are cancelled and 6.6% are rescheduled. For a typical organization the numbers are a little lower: 10% cancelled, 5% rescheduled.
So roughly one meeting in eight doesn’t happen as booked, and one in fifteen moves. That’s the baseline. Everything interesting is in who does it, when, and what they do next.
Most cancellations don’t come from the customer
58% of cancellations come from the host.
That’s the opposite of how the problem is usually framed. The story is always the flaky customer who doesn’t show. In our data, the person cancelling is more often the business.
Some of that is hosts tidying their own calendars rather than genuinely dropping a meeting, so the figure reads harsher than it is. But even allowing for that, more than half of the empty slots are self-inflicted. Before you build a policy to discipline your customers, it’s worth checking your own cancellation rate.
When invitees do cancel, they cancel late
43% of invitee cancellations arrive in the final 24 hours before the meeting. 19% arrive with less than three hours’ notice.
Late notice is the norm, not the exception. Which means a cancellation policy written around “please give us 48 hours” is describing a world that doesn’t exist. Most of what you will actually receive is the three-hour message, and the question is what your business does with three hours.
The further ahead a meeting is booked, the weaker it holds
8.4% of meetings booked for the next 24 hours get cancelled. For meetings booked 30 or more days out, it’s 19.8%. The pattern holds in 85% of the organizations we checked.
And this isn’t a rare corner of the calendar. Half of all meetings are booked six or more days ahead — the median lead time is six days. So a large share of the average calendar is sitting in the window where intent decays.
A calendar that’s full five weeks out is a softer number than it looks.
Half of them don’t cancel at all. They move it.
Here’s the finding that changes what you do about all of this.
When an invitee can’t make the time they booked, 48.8% of the time they reschedule rather than cancel.
Nearly half of what looks like a cancellation problem is actually a scheduling problem, and scheduling problems are fixable. The customer still wants the meeting. They just want it on Thursday.
One reschedule in ten happens after the meeting was already due to start — someone realising at 10:05 that they’ve missed it, and still wanting to talk to you. That’s not a lost customer. That’s a customer with a bad morning.
What a reminder actually does
In the three hours after a 24-hour reminder goes out, invitees cancel 4.7 times as often as they do on bookings with no reminder.
Read quickly, that looks like an argument against reminders. It isn’t.
The reminder doesn’t create the cancellation. It moves it. The person who was going to drop out at 8am on the day now drops out the evening before, because the reminder is the moment they check their calendar and realise. The cancellation was always coming. The reminder decides whether you find out with a day’s notice or twenty minutes.
We can’t tell you that reminders reduce cancellations — our data doesn’t show that, and we’d be making it up if we said so. What it does show is that reminders buy you time. A slot you learn about the night before can be refilled. A slot you learn about at 8:55 cannot.
Hosts who charge upfront see fewer cancellations
For a typical host, 3.9% of paid bookings are cancelled, against 8.3% of free ones. Pooled across all bookings the gap is wider still: 4.4% against 14.5%.
Worth being careful about what this does and doesn’t say. Paid services differ from free ones in a lot of ways beyond the payment — different customers, different intent, different stakes. We can tell you that hosts who charge see about half the cancellations. We can’t tell you that taking payment is what causes it.
What this means for your business
The instinct when cancellations rise is to make cancelling harder: tighter policies, penalties, deposits, friction.
The data points the other way. Nearly half of the people who can’t make it are already trying to keep the meeting, not escape it. Friction aimed at cancellers hits them too, and they’re the ones you want.
The same applies to notice. You aren’t going to talk people into giving you 48 hours. But you can change when you find out. A reminder that surfaces the conflict the night before turns an unrecoverable gap into a slot you can still fill.
And a good share of the problem isn’t customer behaviour at all. It’s yours.
What to do about it
Check your own cancellation rate first. More than half of cancellations come from the host side. Before writing a policy about customers, find out how much of the gap you’re creating.
Make rescheduling more visible. Given the choice, nearly half of invitees move the meeting. If the reschedule link is buried and the cancel button is obvious, you are converting recoverable meetings into lost ones.
Send a meeting reminder, and know what it’s for. A 24-hour reminder might not stop people dropping out, but it will tell you the night before instead of on the morning, and that’s the difference between an empty slot and a refilled one.
Treat long lead times as a risk. A meeting booked 30 days out is more than twice as likely to fall over as one booked for tomorrow. If most of your calendar sits five weeks ahead, plan for the leakage rather than being surprised by it.
Build for short notice. Nearly half of invitee cancellations arrive inside a day, one in five inside three hours. Buffer, waitlist, or an easy route to fill the gap beats a policy that asks for notice you won’t get.
Consider charging for the slot. Hosts who take payment upfront see about half the cancellations — 3.9% against 8.3% for a typical host. That isn't proof payment causes it, but if you already charge afterwards for the appointment, collecting at booking might be a good idea in some cases.
Run your own numbers
What this data does not show
A meeting that isn’t cancelled isn’t necessarily a meeting that happened. We can see cancellations and reschedules. We cannot see no-shows, so our cancellation rates are a floor, not a full account of empty slots.
We also can’t tell you why anybody cancelled. We see the action and the timing, not the reason.
And the reminder finding is about when cancellations arrive, not how many. We have no evidence that reminders reduce cancellations overall, and this piece doesn’t claim any.