Restaurant Overbooking: the Math of the Risk and What to Say When It Goes Wrong

Overbooking has a tempting logic. If 10% of reservations don’t show, why not accept 10% more reservations, so the room is always full? On paper it’s perfect. In reality, it’s a gamble where the stake isn’t your money: it’s the evening of a guest who booked in good faith and ends up without a table. Let’s look at when it makes sense, how much you can really risk, and — above all — how to handle the moment the bet goes wrong.

One caveat: if your goal is to avoid this problem entirely, we have a dedicated piece on how to handle overbooking without losing reservations. Here we start from the assumption that you’re weighing whether and how to use it, and what to do when it breaks.

The math: overbook 60–70% of the no-show rate, not 100%

The most common conceptual error is compensating for no-shows one-to-one. If you have a 10% no-show rate, you accept 10% more reservations. But the no-show rate is an average, not a certainty: some nights it’ll be 3%, others 18%. If you overbook at 100% of the average, every night the no-shows come in under the average you end up with real guests and no tables.

The prudent rule, used by those who overbook methodically, is to compensate for 60–70% of the no-show rate, not 100%. If your Saturday-night slot has a historical no-show of 8%, you can accept roughly 5% more covers on that slot — leaving a margin for the nights when almost everyone turns up.

And here’s the keyword: which slot. Overbooking is never applied across the board. It’s applied only where you have data showing a high, consistent no-show: certain days, certain times, certain booking channels. To know that, you have to measure your no-show rate granularly — exactly as explained in anti no-show strategies. Without per-slot data, overbooking isn’t a strategy: it’s a gamble.

Why it often isn’t worth it at all

It’s worth being honest: for many restaurants, especially small ones, overbooking isn’t worth the risk. The reasons:

The reputational damage is asymmetric. An empty table from a no-show costs you one cover’s revenue. A guest left without a table from overbooking costs you a review, negative word of mouth, and probably that guest forever. The potential loss is far bigger than the gain.

Few tables, thin margin. If you have 12 tables, “5% more” is less than one table: the game isn’t worth the candle. Overbooking only makes sense at volumes high enough to make the average statistically reliable.

Guests treat the booking as a promise. Where guests feel the table is theirs for the evening, blowing up a reservation reads as a serious breach of trust. Overbooking works better where fast turnover is already the norm.

That’s why the smartest move, almost always, is to reduce no-shows at the root — with reconfirmations, reminders, and deposits on at-risk slots — instead of betting on overbooking. If you cut no-shows from 12% to 5%, the problem overbooking was meant to solve almost disappears.

When it goes wrong: the 20-minute rule

Say you practice measured overbooking, and one night everyone shows. Now you have real, booked guests, and fewer tables than you need. How you handle this moment is what separates a small hiccup from a disaster.

The first line of defense is a well-managed wait. This is where the 20-minute rule comes in, a simple principle used in high-turnover venues: if a booked guest has to wait more than 20 minutes because of you, you always send over something on the house — a taste from the kitchen, a glass, a snack. Twenty minutes is the psychological threshold beyond which waiting stops being tolerable and becomes frustration: it’s the moment the guest starts checking the clock and thinking about leaving. A gesture, before that threshold hits, buys back their patience.

Waiting well is a skill in itself, made of honest estimates and an occupied wait: we go deep on it in the psychology of waiting applied to the floor.

What to say to the guest left without a table

If the wait isn’t enough and a booked guest really ends up without a seat, you’re in service recovery territory, and the same rules apply. The script:

1. Immediate, total apologies. The error is yours. “I’m so sorry: we have an organizational problem that’s our fault, not yours.” No technical explanations about overbooking — the guest doesn’t care about your math, they care about their evening.

2. A concrete solution, not a dismissal. The analogy that works is airlines bumping a passenger: they don’t just leave you stranded, they put you on a later flight and give you compensation. In the same way:

“Let me make this right. The aperitif is on us at the bar while I free up the first table that opens — and if you’d rather come back another night, I’ll personally book you the best table with dinner on the house. Your call.”

3. Compensation that actually counts. A glass isn’t enough if the guest crossed town for dinner. The compensation has to be proportionate to the inconvenience: a real gift, a priority booking with an extra. Who can authorize it and up to what amount has to be decided in advance — that’s the logic of the comp policy and authority matrix, because in that moment staff can’t stop to ask permission.

This is, at heart, the service recovery paradox that Will Guidara built a philosophy on: a failure owned and answered generously can leave the guest more loyal than a flawless evening would have.

Mistakes to avoid

Overbooking across the board. Applying it every night, without per-slot data, is the fastest way to end up with real guests and no tables.

Compensating no-shows at 100%. Always leave a margin: the no-show rate is an average, not a guarantee.

Explaining the technique to the guest. “We overbooked counting on no-shows” is exactly what the guest must not hear. The error is yours, full stop.

Letting them wait over 20 minutes without a gesture. It’s the threshold past which you lose the guest’s patience. Always get ahead of it with something on the house.

Using overbooking instead of prevention. If you haven’t first reduced no-shows with reconfirmations, reminders, and deposits, you’re treating the symptom and ignoring the cause.

In summary

Overbooking is a gamble that for many restaurants isn’t worth the risk: the potential loss (a guest, a review, word of mouth) almost always outweighs the gain of one extra table. If you use it, do it methodically — only on slots with high, measured no-shows, compensating for 60–70% of the average, never 100%. And when it goes wrong, handle it as a service recovery: total apologies, the 20-minute rule, real compensation. But the real winning move stays cutting no-shows at the root, so overbooking barely serves a purpose.

The scripts for managing the wait and the guest you need to move are in the Restaurant Floor Scripts Kit below.

Coperti measures your no-show rate by day, slot, and channel, so you know precisely where (and whether) an overbooking margin makes sense — and it manages waits and table reassignment in real time. Explore the features or tell us about your venue for a demo.