Every salon owner reaches the same fork. The no-shows are costing real money, somebody suggests deposits, and the room splits between "obviously" and "our clients would hate it."
Both sides are arguing from a real position, and neither can win, because the argument turns on a number nobody has.
The number you have, and the one you don't
A no-show is visible. It has a name on it, a time, and an empty chair you can point at. You can count them, and once you count them you can multiply, and the figure is usually bad enough to make deposits sound obvious.
The client who decided not to book once you mentioned a deposit is invisible. They didn't cancel — they never appeared. Nothing in your system records them, so nothing in your reasoning includes them.
So every deposit debate is one measured number against one unmeasured one. That's why it never resolves, and why it usually gets settled by whoever feels most strongly rather than by arithmetic.
What to try before deposits
Deposits are the biggest lever and the one with the worst side effects. There are cheaper ones, and they're worth exhausting first.
Make the reminder something they answer. A text into the void is not a reminder; it's a notification. A reminder they reply to — even just "Y" — converts a portion of would-be no-shows into either an arrival or a cancellation early enough to matter. The difference between those two outcomes is the whole game.
Make cancelling easy at the hour people actually decide. This is the one most salons get wrong, and it isn't a policy problem. People decide to cancel in the evening, and most salons cannot be reached in the evening. A client who tries at 8pm, gets voicemail, and gives up becomes a no-show in your records — but they weren't a no-show. They were an unanswered phone.
That distinction matters commercially, not just semantically. A slot released at 8pm the night before can be refilled. The same slot released when you open at 9am usually cannot.
Confirm the day before, not the week before. A reminder seven days out is filed and forgotten. Twenty-four hours out is when someone still has time to move it rather than abandon it.
Look at who. No-show rates are rarely spread evenly. Usually a small group of clients account for most of them, and a quiet word or a card on file for those specific people solves more than a policy applied to everyone who has never let you down.
If you do want deposits, ask a better question
Not "do deposits reduce no-shows" — they generally do. The question is whether they reduce them by more than they cost you in bookings that never happen.
That has an answer, and it's one line of arithmetic. The bookings you lose cost you the full ticket. The no-shows you prevent were costing you the ticket minus whatever deposit you'd have kept. Set those equal and you get the deterrence rate at which deposits stop being worth it.
With a 12% no-show rate, a $75 average ticket, a $20 deposit, and no-shows falling to 4% once you take them, that line sits near 9%. So if fewer than about nine in a hundred clients would book somewhere else rather than pay a deposit, deposits pay for themselves. More than that, and they don't.
You still don't know your real deterrence rate. But you now know what it has to beat, which is a much easier thing to have an opinion about than a number in the abstract. The calculator does this for your numbers.
How to find out for real, cheaply
If your answer flips between plausible guesses — if 5% and 15% give you different conclusions — you're guessing, not deciding. The way out isn't more debate.
Put a deposit on one service for one month. Pick something with a high no-show rate and a long duration, where the loss hurts most. Watch two numbers: whether no-shows on that service fall, and whether bookings for it fall. You'll know inside a month, and you'll have risked one service instead of your whole book.
That's a cheaper experiment than the argument is, and considerably cheaper than being wrong in either direction for a year.
The part nobody counts
Before any of this, it's worth finding out how much of your no-show rate is really an unanswered phone.
Go through last month's no-shows and check whether each one tried to reach you. If you can't tell — if there's no record of calls that rang out — that's the finding. A meaningful share of what gets recorded as "didn't show" is "tried to tell you and couldn't."
That share doesn't need a policy. It needs someone to pick up.