Somebody booked. The confirmation went out, the slot came off your calendar, and at two o'clock you sat in an empty call. The instinct within the hour is to add another reminder.
Reminders are the smallest lever in this problem and everybody has already pulled it. A fourth nudge on the morning of the call is not what is missing. Something upstream decided this days ago, usually at the moment the booking was made.
The uncomfortable version: most no-shows are booked, not lost. The reminder is only where you found out.
A no-show is a qualification failure with a later date
Look at what a frictionless booking asks of a stranger. A grid, a time, a name and an email address. Twenty seconds, nothing given up. Excellent as a conversion flow, poor as a predictor of attendance, because nothing in it required the person to decide anything.
Attendance is not produced by the reminder. It is produced by commitment, and commitment is made of the small costs somebody accepts on the way in. Answering three questions honestly is a cost. Choosing a budget band is a cost. Neither is large, and size is not the point. The person spent something, and people turn up to things they have already spent something on.
This is the argument in qualifying leads before they book arriving from the opposite direction. That post is about meetings you should never have taken. This one is about meetings you never got to take. They share a cause: the page asked for nothing and got it.
So when the show rate slips, look at what your page asks of a stranger before you touch your reminder cadence. If the answer is a name and an email address, the number is reporting honestly on what those bookings were worth.
Friction trades bookings for attendance
Here is the part most writing on this leaves out, because it complicates the advice. Adding steps to a booking flow does lower the number of bookings. There is no version where you get both.
Some arithmetic to make the shape visible. I made these numbers up. They are illustrative, and the only ones that matter are yours.
- Open page. Forty bookings in a month. Six in ten turn up: 24 meetings held, 16 slots spent on people who never arrive.
- Gated page. The questions cut bookings by a third, to 27. Eight and a half in ten turn up: 23 meetings held, 4 slots lost.
One meeting fewer, twelve slots back. A no-show spends a slot exactly as an attended meeting does, and nothing in a bookings number says which of the two you bought.
Now run the same percentages on twelve enquiries a month rather than forty. Twelve bookings at six in ten is seven meetings. Gate it down to eight bookings at eight and a half in ten and you still hold seven. Three slots recovered that you had no use for, and nobody new spoken to. The friction bought nothing, because calendar time was never the constraint.
The variable nobody chose: how far ahead people can book
Ask a team what their booking horizon is set to and most cannot tell you. It arrived with the tool and stayed. After qualification it is the largest thing here you can move this afternoon.
Intent decays. Somebody filling in your form today has a problem today: something broke, a contract is up for renewal, a director asked a question this morning. Three weeks later that problem has been solved, worked around or absorbed into how things are. They are not being rude when they miss the call. They are a different person from the one who booked it.
A call booked for tomorrow rides the same wave of intent that produced it. A call booked for three weeks on Thursday has to survive it. No reminder changes which of the two you are holding.
So cap it. A fortnight covers almost every inbound sales call, and a week is plenty if you have the volume to fill one. The objection is fair: a shorter horizon means some people cannot find a time and go elsewhere. That is a real loss, so move the number one step at a time rather than straight to three days on a hunch.
In Setupp both ends of that range sit on the event type: how far ahead the calendar opens, which ships at thirty days, and how much notice you need before the first slot. Thirty days is a default, not a recommendation. If you have never changed it, you are running somebody else's opinion about your pipeline.
One reminder does the work. The calendar does the rest.
Reminders in their proper place, then: useful, small, and not the thing that was wrong. The one that matters is not an email at all. It is the calendar invite. An email lands in an inbox, where it sits under forty other people's requests by lunchtime. An accepted invite lands in the thing somebody opens to find out what they are doing today.
Which makes the useful questions mechanical rather than editorial. Does a real invitation go out from a calendar system the booker's own calendar recognises, and is that person an attendee with a Yes and a No to press? If either answer is no, your emails are doing a job the calendar was built for.
With the invite doing its job, one email the day before is enough, and it earns its place by carrying the join link and the time in the reader's own zone. A second reminder is a small addition. A fourth is a message about you rather than about the meeting, and somebody counting six touches for a thirty-minute call learns how badly you need it.
Make cancelling easy on purpose
This one gets argued with, so plainly: a cancellation is a good outcome. Not as good as a meeting. Far better than a no-show, and that distance is larger than most teams behave as though it is.
- You get the slot back, usually in time to give it to somebody else. A no-show returns it at the moment it becomes worthless.
- You learn something. A cancellation is an answer, often with a reason attached. Silence at two o'clock is not data about anything.
- The relationship survives. Somebody who cancelled can book again without embarrassment. Somebody who vanished feels they cannot, and does not.
Hiding the cancel link does not prevent cancellations. It converts them into no-shows. Somebody who has decided not to come has already decided, and the only question is whether telling you is cheap or awkward. Make it a two-click job and part of your no-show rate becomes a Tuesday morning email you can act on. It will look as though the number got worse. What happened is that a real number came out of hiding.
A cancellation is information. A no-show is the same decision with the information withheld.
The version that saves the deal is the reschedule. If the link that cancels also offers a new time in the same flow, some of the people who were leaving take that instead.
Measure the pair, not the rate
Show rate on its own is the easiest metric in sales to improve, and the way to improve it is to book fewer meetings. Any gate, any horizon cut, any extra question moves it up. A team can drive show rate to nine in ten while quietly starving, and the dashboard reads as a success story throughout.
So carry two numbers together and never quote one without the other.
- Show rate. Of the meetings booked, how many happened.
- Meetings that went somewhere. Of the meetings that happened, how many produced a next step you would have written down anyway.
Both rising is the good outcome and the common one, because the same causes drive them. Show rate up with the second number flat means you filtered on willingness rather than on fit. The second number up with show rate down points at distance rather than qualification, so check the gap between booking and meeting first.
Record why each missed meeting was missed, too: cancelled, moved, or vanished without a word. Three different problems, and only the last one is about reminders.
Where Setupp fits, briefly, because the order matters more than the tool. The qualification gate is the upstream lever and the only one here that changes who books at all: the questions render before the calendar, the rules run on the server, and somebody who does not clear them gets a message you wrote rather than a slot. Turning people away is a paid feature, as pricing sets out.
The reminder is one email, a day ahead, sent once. Not once by convention: the email log holds a unique index on the booking, the template and the recipient, so an overlapping sweep is refused by the database rather than delivered as a second copy. Every booking also writes a real invitation to a connected calendar, with the booker and every host on it as full attendees genuinely asked to respond, and carries a cancel and a reschedule link in the body of the confirmation rather than in grey type at the bottom.
Then the loop closes, which is the part a scheduler normally cannot do. The host marks each meeting completed or a no-show, so the dashboard can report attendance against what the Setupp Score predicted, band by band. It refuses to show a rate until eight outcomes are in, because one absence out of three reads as 67 per cent and teaches you to distrust the panel before it has anything true to say.
If more than one of you sits on these calls, what lands on each person's calendar is its own question, covered in round-robin and collective scheduling. Setupp is that order in one place. None of it manufactures attendance out of nothing. It moves the decision to where it is cheap: before the slot is taken, rather than at two o'clock on a Thursday when the half hour is gone anyway.