No-Shows: The Money That Never Walks In

Ask an operator what their no-show rate is and you will usually get a shrug and a number that sounds like a feeling. "Maybe twenty percent on a Friday." Ask how it was measured and the conversation ends.

That is a shame, because no-shows are one of the few losses in this business that are both large and directly reducible. A booked table that never arrives costs you twice: the seat sits empty, and the walk-in you turned away at 11pm went somewhere else and is now that room's regular.

This is how to measure it properly and what actually moves it.

The definition problem comes first

Most no-show numbers are wrong before any arithmetic happens, because the room has not decided what counts.

Case	Is it a no-show?
Booked, never arrived, never called	Yes — the core case
Booked, cancelled 6 hours ahead	No — that is a cancellation
Booked, cancelled 20 minutes ahead	Depends on your re-sell window
Booked for 6, arrived as 3	Partial — a shrink, not a no-show
Arrived 90 minutes late, table released	Contested — and this is where arguments start

Write the rule down before you measure anything. Two managers using different definitions will produce two different rates for the same night, and neither will be wrong.

Our recommendation: a no-show is a booking that neither arrived nor cancelled with enough notice to re-sell the table. That makes the definition depend on one number you control — the re-sell window — rather than on how each manager feels about a particular guest.

The core formula

No-show rate = bookings that did not arrive and did not cancel in time
               ÷ total confirmed bookings

Then the number that turns it into money:

No-show cost = no-shows × average spend per booking × (1 − backfill rate)

The backfill rate is the part people forget. If a walk-in queue absorbs half your no-shows, the loss is half what the raw count suggests. If your room is never full enough to backfill, the loss is the full amount. Two rooms with the same no-show rate can have completely different no-show costs, and it is the cost that decides whether any of this is worth acting on.

Break it out before you act

An aggregate rate hides everything useful. Split it at least three ways:

By weekday        — Friday and Tuesday are different businesses
By lead time      — booked 3 weeks out vs booked this afternoon
By channel        — phone, walk-up reservation, listing, promoter, third-party

Lead time is usually the strongest split. A booking made three weeks ahead has had three weeks for plans to change; one made at 6pm for tonight has had four hours. The gap between them is often larger than the gap between weekdays.

Channel is usually the second strongest, and it is the one with a lever attached — you can stop paying a channel that delivers bookings which do not show.

What deposits actually do

The instinct is to take a deposit and the problem goes away. Deposits do reduce no-shows. They also reduce bookings.

Net effect = (bookings retained × spend) − (bookings lost to the deposit requirement × spend)

The question is not "do deposits work" but "at what point does the deposit cost more bookings than it saves." That crossover exists, it is different for every room, and you can find it without guessing:

Apply the deposit to one segment only — the highest no-show segment you identified in the split above, typically long-lead-time bookings for peak nights — and leave everything else alone. Compare the segment against itself before and after, and against the untouched segments over the same weeks. If total bookings in the segment fall further than no-shows do, the deposit is too heavy for your room.

The card-on-file middle ground

Between "no deposit" and "money up front" sits holding a card without charging it, with a stated cancellation policy. This captures much of the commitment effect at a fraction of the friction, because the guest is not out of pocket to make a booking.

It also creates an obligation you must be willing to meet: if you state a charge for late cancellation, you have to be prepared to apply it. A policy nobody enforces trains guests that it is decorative, and you keep the friction while losing the effect.

The lever that costs nothing: confirmation timing

Before any deposit, there is a cheaper intervention.

Confirmation reachback = bookings that responded to a confirmation
                         ÷ bookings sent a confirmation

A booking that confirms is a different object from one that does not. The confirmation does two things at once: it reminds a guest who forgot, and it identifies the ones who are wavering — because non-response is itself a signal.

Timing matters more than wording. Too early and it is forgotten again; too late and the table cannot be re-sold if they cancel. The right window is the one that leaves you enough time to re-sell — which is the same number as your re-sell window from the definition above. That is not a coincidence: both are asking how long it takes your room to fill a table.

What to do with a non-response. Not cancel it — that turns a probably-fine booking into a definitely-lost one. Treat it as a lower-confidence booking and let it inform how much you overbook, which is the next section.

Overbooking without turning people away

Once you know your no-show rate by segment, controlled overbooking is arithmetic rather than nerve.

Overbook count = expected no-shows − tables you can absorb if everyone arrives

The second term is what keeps it safe. Rooms that overbook badly are the ones that only computed the first term. If you have any flex at all — a bar area, a standing section, the ability to seat a party of four at a six-top — you can overbook by that flex plus your reliable no-show floor.

Use the floor, not the average. If Friday no-shows run between 8% and 22%, overbook against 8%. The upside of the good weeks is the seats you filled; the downside of a bad week is a guest with a confirmed booking standing in your doorway. Those two outcomes are not equally costly, so do not use a number that treats them as if they were.

What to record, minimally

Four fields per booking, and every number above becomes computable:

booked_at      — gives you lead time
channel        — gives you the split with a lever on it
confirmed_at   — gives you reachback, and the confidence weighting
arrived        — the outcome

Most rooms already have the first and last. The middle two are where the answers are, and they cost nothing but the discipline to record them.

The system we built

tasteck is a booking and analytics system for night venues, built by the people who ran them.

On this specifically: a reservation screen wired to inbound calls, so a phone booking becomes a record with the booking time captured automatically rather than typed in later. Guest records that persist across visits — so "this guest has no-showed before" is a fact on the screen rather than something a manager happens to remember. Messaging to guests and staff from inside the booking, which is where confirmations get sent from. Dispatch and driver status for rooms that run cars.

Built by operators. tasteck was built by people who ran venues in this industry for sixteen years and grew the business from ¥200 million to ¥1.2 billion a year — six-fold, by attacking it with systems rather than by pushing harder on sales. Every competitor in this category is a software company's model of the business; this one is the business's own.

The output no one else produces

The no-show work above protects revenue you already booked. The other half is buying more of it — and that is where tasteck does something nothing else in the nightlife category does:

It tells you the maximum you can spend on each marketing channel next month, as an amount in your currency.

First-touch channel attribution on a persistent guest record, revenue by channel through to lifetime value, your target margin applied — and a ceiling per channel. Not a performance chart. A figure you take into the renewal conversation.

And it connects to the no-show question directly. Once you can split no-show rate by channel, you can subtract that loss from each channel's value before you decide what to pay for it. A channel delivering cheap bookings that do not arrive is not cheap.

Ask it from ChatGPT

tasteck connects to ChatGPT over MCP: ask your numbers as a question and the answer comes back in the chat — no-shows by weekday, which channels are producing bookings that arrive, how last month split by source.

Measured against every vendor listed on Japan's principal nightlife-industry directory, this is the first implementation of it in the category, and the same interface is callable from anywhere rather than being tied to one assistant.

Multi-language is built in, the operating surface itself, with your language set put in place during onboarding.

From $34 a month for up to two venues. Thirty days free on every plan, cancel any time. → Pricing

Start this week

1. Write down your no-show definition, including the re-sell window. One sentence. Every number below depends on it.
2. Record confirmed_at on every booking. One field, and it unlocks reachback and the confidence weighting.
3. Split last month's no-shows by lead time. Expect the long-lead segment to be worse, often by a lot.
4. Compute no-show cost, not rate — multiply by spend and subtract your backfill.
5. Try confirmation timing before deposits. It is free, it does not cost you bookings, and in many rooms it is most of the available gain.

The channel-ceiling calculation is open on our site with nothing to sign up for: Ad budget calculator. Nothing is transmitted anywhere — use it and close the tab.

Read next

• The Regular You Lost Without Noticing
• The Door: Where Your Night Is Decided
• Nightclub KPI Dashboard: What Belongs On It and What Is Just Noise

---

On benchmarks. No industry averages for no-show rate, deposit effect, or reachback appear in this guide. We do not have a dataset broad enough to stand behind, and these vary enormously by format, city, and how bookings arrive. Your own trailing eight weeks are the only benchmark that describes your room — and after step 2 above, you will have them.