Why the figures never agree
Two no-show figures can only be compared if they answer the same questions. Each study makes its own choices.
- What is counted? Strict no-shows (nobody comes, nobody calls), or no-shows and last-minute cancellations together. For the dining room the effect is the same, an empty table. For the statistic, the gap can be large.
- Out of what? Bookings or covers. A party of eight that fails to show weighs the same as a table for two in the first calculation, and four times as much in the second.
- Over what period? A yearly average smooths everything out. A daily rate shows you Saturday night.
- Who recorded it? A figure drawn from booking software only counts the no-shows restaurants bothered to mark. An absence nobody logged does not exist in the average.
What the verifiable studies say
Academic studies of restaurant no-shows are rare and often cover a single venue. Their strength: you know exactly what was measured.
A hotel restaurant in Hong Kong
Tony Tse, of the Hong Kong Polytechnic University, and Y. T. Poon, of Iowa State University, analysed the bookings of a hotel restaurant over 27 months, from April 2012 to June 2014: 749 lunches and 749 dinners, after removing the weeks when a deposit was required. They count no-shows and late cancellations together. The result: a probability between 9.07% and 12.95% depending on the day, lowest on Thursday and highest on Saturday, at lunch and dinner alike.
So the rate shifts markedly from one day to the next in the same restaurant, which a single average hides.
A restaurant in Taiwan
C. I. Chiang, in a study published in 2021 in the Journal of Hospitality & Tourism Research, used data from one Taiwanese restaurant to illustrate an overbooking model. That restaurant’s daily no-show rate ranged from 11% to 16%. Again: one venue, and a figure that depends on the day. Whether you should overbook your own room is another question, covered in our article on overbooking a restaurant.
| Study | What is counted | Setting | Result |
|---|---|---|---|
| Tse and Poon, 2017 | No-shows and late cancellations, per booking | One hotel restaurant, Hong Kong, 2012-2014 | 9.07% to 12.95% depending on the day |
| Chiang, 2021 | No-shows, per day | One restaurant, Taiwan | 11% to 16% depending on the day |
The figures that travel without a source
“One booking in seven”, around 14%, is credited to a British study and repeated from blog to blog, in English and in French. We could not find the original report or its method: it is a figure that circulates without a verifiable source.
Averages of 2 or 3% usually come from booking platforms, published by the companies that run them. They cover many restaurants, but they only count the no-shows marked as such in the software, mix restaurants that take a card guarantee with those that do not, and are not reviewed by independent researchers.
None of these figures tells you what happens in your dining room.
Working out your no-show rate
The formula
No-show rate = bookings that did not turn up without notice ÷ bookings expected at the service × 100.
Expected bookings are the ones still valid at service time: take out the cancellations made in good time, which are not no-shows. Count late cancellations separately, for example those received less than two hours before the booked time: they also leave a table that is hard to resell.
The period
Take at least eight weeks and split your services: lunch and dinner, weekdays and weekends. That is where the differences hide, as the Hong Kong study shows. This rate is one of the five booking numbers to look at in your year in review.
An example
As an illustration: over eight weeks, your dinner service expected 400 bookings. 18 did not turn up without notice, and 9 cancelled less than two hours ahead. Your no-show rate is 18 ÷ 400 = 4.5%. Add the late cancellations and the share of lost tables rises to 27 ÷ 400 = 6.75%. Run the numbers again in covers: if the missing guests were mostly large parties, the figure climbs further.
Log every absence
The calculation only works if each absence is recorded on the day, on the booking itself, not from memory at the end of the month. With gonnabook.it, the “No-show” status is set on the booking from the service list, and the booking details then show the guest’s visits and no-shows, counted automatically from those statuses.
What your number tells you
What matters is where the rate concentrates.
- One day or service stands out? Start there: a reminder the day before, confirming large parties, a card guarantee for that evening only.
- The no-shows are mostly bookings made weeks ahead? Forgetting is the likely cause, and a well-timed reminder deals with it.
- The same names keep coming back? Record the facts, without judgement, and adapt your terms for those guests.
- Changing something? Keep measuring: the numbers will tell you whether it worked.
The levers themselves are covered in our guide to reducing restaurant no-shows, and the rules to announce to guests in your written booking policy.
Frequently asked questions about no-show rates
What is the average restaurant no-show rate?
There is no single reliable average. The academic studies available each cover one restaurant: roughly 9% to 13% in a Hong Kong hotel restaurant, late cancellations included, and 11% to 16% per day in a Taiwanese restaurant. The averages of a few per cent published by booking platforms only count the no-shows marked in their software, so they are not directly comparable.
How do I calculate my restaurant’s no-show rate?
Divide the number of bookings that did not turn up without notice by the number of bookings still expected at service time, then multiply by 100. Leave out cancellations made in good time, track late cancellations separately, and run the calculation over at least eight weeks, splitting lunch and dinner, weekdays and weekends.
Is a last-minute cancellation a no-show?
Not strictly: the guest did let you know. But for the dining room the effect is similar, because a table is hard to resell so close to service. Some studies count them together. The simplest approach is to track two numbers, no-shows and late cancellations, with a cut-off you choose, for example two hours before the booked time.
Sources
- Tony S. M. Tse and Y. T. Poon, Modeling no-shows, cancellations, overbooking, and walk-ins in restaurant revenue management, Journal of Foodservice Business Research, vol. 20, no. 2, 2017, Hong Kong Polytechnic University repository, accessed 18 November 2026.
- C. I. Chiang, Overbooking as a Means to Manage Restaurant No-Shows and Cancellations: A Novel Model Extension, Journal of Hospitality & Tourism Research, 2021, abstract, accessed 18 November 2026.
- The “one booking in seven” figure is quoted as a figure that circulates without a verifiable source.
- The worked example (400 bookings, 18 no-shows, 9 late cancellations) is an illustration, not a statistic.