The Deposit You Never Take

A table for eight on a Saturday. Confirmed twice. Nobody arrives.

You held the best section of the room for two hours and billed nothing, and the reason you did not take a deposit is that asking felt like it would have lost the booking.

What not asking costs

The lost table is the obvious part and the smallest part.

The revenue of the slot. Your highest-value hours, held and unsold.

The turn-aways. ⚠️ The expensive one. You declined other parties for that table. Those are real bookings you refused in favour of one that never came.

The staffing. You rostered for the covers you expected. That decision was made days earlier and cannot be undone at 21:00.

What asking costs

Some bookings, genuinely. Far fewer than people assume, and the ones you lose are disproportionately the ones that were going to no-show.

That is the whole argument. A deposit request is a filter, and what it filters out is precisely the segment that was never firm.

⚠️ A booking that evaporates when you ask for £20 was not a booking. It was an intention, and you were carrying all of its risk.

Which bookings deserve one

Not all of them. Asking universally is where this goes wrong — it adds friction to your reliable regulars for no gain.

Large parties. The blast radius is proportional to size.

Peak slots. Where the opportunity cost of an empty table is highest.

First-time bookers on a busy night. ⚠️ No history, highest-value slot — this is the exact intersection where the risk sits.

Anything with a held resource. A specific person requested, a room set aside, stock bought in.

Not: known regulars, off-peak, small parties. The filter is not free, so point it at the risk.

Making it uncontroversial

The wording does most of the work.

Frame it as confirmation, not distrust. "To confirm the table" reads differently from "in case you do not turn up", and they are the same sentence.

Make it redeemable. ⚠️ A deposit that comes off the bill is not a charge — it is a prepayment, and guests understand the difference instantly.

State the cancellation window plainly. Forty-eight hours, twenty-four, whatever you choose. Ambiguity here is what generates the arguments, not the policy itself.

Same rule for everyone in the category. Discretionary application is worse than no policy — it produces the one conversation you were trying to avoid.

The part that actually fails

Not the guest reaction. The taking of it.

A deposit policy that requires someone to phone back with a card number collapses within a month, because it doubles the work of every booking and the person taking calls will quietly stop.

⚠️ If it is not collectable in the same interaction as the booking, it will not be collected. That is an operational constraint, not a discipline problem, and no amount of reminding fixes it.

What to watch

No-show rate on deposit-taking bookings versus the rest. The number that settles the internal argument.

Bookings lost at the deposit request. Count them honestly — this is the cost side, and it is usually much smaller than the fear of it.

Deposits taken versus deposits forfeited. ⚠️ If you are forfeiting a lot, the policy is working as insurance. If you forfeit almost none, it is working as a filter. Both are the policy working, and they look completely different in the accounts.

Where the record has to sit

A deposit is only useful if the person on the door knows it was taken. If that lives in a separate payment system from the booking, the guest arrives having prepaid and is asked to pay again — which costs more goodwill than never having asked.

tasteck keeps payment status on the booking record itself, so the deposit is visible wherever the booking is.

You are not avoiding an awkward conversation by not asking. You are having a different one, later, with the party you turned away.