Every restaurant owner knows the feeling: four no-shows on a Friday night, and there was a waitlist for that exact table. The seats sat empty and so did the revenue. Overbooking fixes that — but almost nobody in hospitality does it on purpose, with a number behind it instead of a guess.
Airlines and hotels have overbooked systematically for more than fifty years. They know from experience that a predictable share of bookings never shows up, and they deliberately sell a few more seats or rooms than they have — enough to fill the gap the no-shows leave, not so many that they routinely have to turn people away. It's called revenue management, and it's taught in every hospitality programme.
Restaurants almost never do the equivalent, despite carrying the same no-show problem. Part of the reason is that nobody has ever explained it without reaching for a statistics textbook. The other part is that the first time it goes wrong — a table of six that all actually show up while you're already full — feels like a disaster, and nobody counts the nights it quietly worked in your favour.
This isn't an argument for booking recklessly. It's a way to put a number on what you're probably already doing by feel: "we'll take two more, someone always cancels." Seven numbers, in order, and a calculator at the end that combines them with your own seat count and your own no-show rate.
Everything runs in your own browser: nothing is sent anywhere and nothing is saved. This is a simplified model, not a guarantee — your own history, your own guests and your own kitchen are the final word.
The ultimate guide Restaurant Reservations Guide: 6 Steps to a Full House How to manage restaurant reservations: cutting no-shows, refilling cancellations, pacing the rush and filling quiet nights, in one guide. Open the guideThe 7 numbers, in order
Each number builds on the last. Start with your own no-show rate — without it, the rest of this guide is just theory.
1. Your own no-show rate
This is the one number on this page you can't calculate — you have to count it. Pull your reservations from the past two to three months, per service, and count how many actually showed up against how many were booked.
Two mistakes happen here most often. First: a reservation that was cancelled ahead of time is not a no-show — that's exactly why you got the cancellation, and you may have been able to resell the table. Count only the people who were booked, never showed and never said a word. Second: measure per service, not per month. A monthly average of 10% hides a Friday dinner running at 4% and a Tuesday lunch running at 22% — and it's Friday dinner where the margin actually pays off.
No history yet? Start with an estimate of 10 to 15% for phone and in-person bookings, and start counting from today — after three weeks you'll have your own figure instead of one borrowed from a blog post.
2. Why it isn't one number
A 12% no-show rate sounds like a single figure, but it's an average of very different behaviour. A phone reservation made two weeks out by someone who already knows the place rarely no-shows. An online reservation for the same night, made by someone who still had three other restaurant tabs open, no-shows a lot more often.
Below is an illustrative example for one busy Friday dinner service — not a universal industry statistic, but the pattern reservation platforms and restaurants keep seeing: the later and less personal the channel, the higher the no-show rate tends to run. Large groups carry a separate risk on top: the more people have to agree, together, to actually turn up, the higher the odds one drops and the rest follow.
The point isn't that you need a separate formula per channel — that would overcomplicate this guide. The point is that the one no-show rate from step 1 is an average, and it's worth knowing which slice of your bookings is dragging it up. If you do overbook, do it against the channels causing that average, not against your regulars.
An illustrative example for one busy Friday dinner service — not a universal percentage, but the pattern reservation platforms keep reporting: the later and less personal the channel, the higher the no-show rate tends to run.
Large groups no-show more often for a different reason than late online bookings: the more people who have to agree, together, to turn up, the higher the odds one drop-out takes the rest with them. Know your own channels, and overbook against the ones dragging your average up.
3. The asymmetric cost: an empty table vs. a turned-away guest
This is the reason most restaurants don't overbook, even though it's essentially free money most nights. An empty table from a no-show feels like bad luck — annoying, but nobody calls you about it. Turning a guest away because you overbooked and everyone actually showed up feels like a mistake — your mistake, in real time, with a frustrated party standing in front of you.
Those two costs aren't symmetric, and that's exactly why they weigh so much heavier psychologically than they should financially. An empty table costs you that one cover's revenue, for one night. A turned-away guest costs you that same lost revenue plus an unhappy guest, a possible bad review, and the conversation you have to have yourself, at the table or on the phone.
That's exactly why the safety margin in step 5 needs to be wider than a pure expected-value calculation would suggest — not because the maths demands it, but because the two outcomes don't feel equal. Price the asymmetry in rather than ignoring it, and overbooking becomes a deliberate trade-off instead of a gamble that occasionally backfires.
4. The overbooking formula
At its core it's simple: if a predictable share of your reservations don't show up, you can plan for that share ahead of time instead of just eating the loss. Run 40 covers and expect 12% of bookings to no-show, and you could in theory take up to roughly five reservations beyond your seat count without averaging more guests than you have tables.
That figure — the raw margin — is just your capacity multiplied by your no-show rate, divided by the share that does show up. It's the expected value: the number of extra reservations that, on average, exactly cancels out your expected no-shows. No more, no less.
The problem with the raw margin on its own is that "on average" isn't the same as "every night". Some nights everyone shows up anyway. The raw margin tells you the average; it says nothing about how often you'll actually land over capacity. That's what step 5 is for.
5. The safety margin — why a restaurant needs a wider buffer than an airline
An airline that overbooks often books close to that raw margin — the point where the expected number of arriving passengers equals the number of seats. That sounds sensible until you realise what it means: at that margin, the odds that too many people actually show up run close to a coin flip. The airline accepts that because it has a compensation system standing by: a volunteer who takes a voucher and the next flight, an upgrade offered to whoever stays behind. The oversell moment is priced in and solved before it happens.
Your restaurant doesn't have that system. There's no volunteer trading a voucher for a later booking. There's a party standing at your door, and an apology nobody enjoys giving or receiving. That's why this guide works with three caution levels, all of which book fewer reservations than the pure raw margin would allow: Conservative trims the largest safety buffer off the raw margin, Balanced a smaller one, and Aggressive books right up to the raw margin — with the coin-flip risk that comes with it.
At 40 covers and a 12% no-show rate, that works out to: Conservative lands on 43 reservations, Balanced on 44, Aggressive on 45 — the full raw margin from step 4. The gap between the three is small in reservation count, but large in how often you'll have to disappoint a guest. Below is where the three levels land on the line relative to your capacity.
A restaurant with 40 covers and a 12% no-show rate. Each level trims a different safety buffer off the same raw margin — which is why three levels produce three different "safe" reservation counts.
Aggressive books right up to the full raw margin, which works out to roughly a coin-flip chance of a night where you have to make someone wait. Conservative leaves the most room — which is exactly why step 7 recommends starting there.
6. Your safety net for when it goes wrong anyway
Even at Conservative, there will be a night when everyone shows up. Overbooking without a safety net is taking the risk without the cover; the safety net is what makes overbooking workable rather than nerve-wracking.
Four things every restaurant that overbooks keeps standing by. A walk-in list: guests who arrive without a reservation are first asked to wait ten to fifteen minutes at the bar rather than seated immediately, which buys you slack on a busy night. Bar seating counted as a real table for two: a couple who's early for their booked table can perfectly well start at the bar. A standby call: ring your two most likely no-shows of the night about thirty minutes ahead to confirm — whoever doesn't answer, or hesitates, gives you a more realistic picture before it becomes a problem instead of after.
And if it does go wrong anyway: a drink on the house at the bar while the next table is turned costs a fraction of what a turned-away guest costs you in reputation, and it signals the house has the situation under control — even when, for ten minutes, it isn't quite.
7. The rollout plan
Don't start at Aggressive, even if the calculator below shows it "checks out" on paper. The advice already elsewhere on this site is the right starting point: if your no-show rate runs 10%, start with roughly 5% overbooking — in practice closer to Conservative than Balanced, until you've built up a few weeks of your own experience.
Track two numbers weekly: how many reservations you overbooked, and how many times you actually had to make someone wait at the bar or ask them to come back later. If that second number stays at zero for several weeks running, you're leaving margin on the table and can nudge towards Balanced. If it's happening several times a week, you're running more aggressively than your safety net can absorb, and it's time to pull back a level.
This isn't a setting you pick once and forget. Your no-show rate shifts with the season, with the weather, with who happens to be on the phone confirming bookings that week. Recalculate your rate every quarter, and adjust your caution level the way you'd adjust a staff schedule.
Work out your own safe margin
Enter your own capacity and no-show rate — the fields are seeded with the example from step 5, so overwrite them with your own figures from step 1. Then pick a caution level; everything below recalculates instantly.
The calculator returns how many reservations you can safely take, how many that is beyond capacity, how many no-shows you can expect at that count, and how likely it is that you'll still have to make a guest wait or turn one away.
Overbooking calculator
Your own capacity, your own no-show rate, and the reservation count that goes with it.
—
This is a simplified model built on your own two numbers, not a guarantee. Everything runs in your browser; nothing is sent or saved. Channel, day, season and weather all shift your real no-show rate — see step 2.
Two things to keep in mind when you read the result. Margin and risk aren't two separate dials you can optimise independently — they're two sides of the trade-off from step 3: every extra reservation lowers the odds of an empty table and raises the odds of a turned-away guest, at the same time.
And the risk percentage is an average across many nights, not a forecast for tonight. Most nights you'll land exactly right or just under capacity; it's the rare night where everything lines up that step 6's safety net is there for.
Overbooking isn't a gamble, it's a calculation
Most restaurant owners already overbook informally, without ever calling it that — "we'll take two more, someone always drops out". All this guide adds is a number behind that instinct: your own no-show rate, your own capacity, and a caution level that matches how solid your safety net actually is.
Start conservative, track weekly how often you actually have to make someone wait, and build up from there. That's the same advice already given in reducing no-shows, and the two guides deliberately don't contradict each other: cutting no-shows at the source remains the first move, overbooking catches what's left over.
If you'd rather take no gamble at all on a turned-away guest, pair overbooking with a waitlist: seat guests only once space has genuinely opened up, instead of betting in advance that it will. The two together give you the best of both — the revenue of overbooking, the certainty of a waitlist.