Restaurant no-show rate: the Resos No-Show Index 2026

According to Resos data covering 3,768,761 reservations across 2,417 restaurants, the average recorded restaurant no-show rate is 2.33%. That figure comes from 87,953 recorded no-shows between August 2025 and July 2026, representing 283,728 booked covers that never arrived.

This is the Resos No-Show Index: a recurring benchmark built from real reservation outcomes in our booking system, not from surveys. Every rate in the tables below states its denominators, and the full method is documented further down. The index is updated with each new edition, always at this address.

Why 2.33% is bigger than it sounds

A low percentage hides a high cost, for three reasons.

It compounds nightly. 2.33% of everything means the average restaurant in this dataset recorded about 36 no-shows a year, walking away from roughly 117 booked covers. A busy venue in the 500 to 999 bookings-a-month bracket runs at 3.17%, well above the average.

No-shows are only the visible half. This dataset also contains 566,049 cancellations, and 56% of them arrived within 24 hours of the booking, or after it had already started. A table released that late is nearly as hard to resell as a no-show. Counting both, restaurants in this dataset lost about 405,000 tables in a year to guests who booked and did not come, roughly 168 per restaurant.

They hit the most valuable service. The no-show rate at dinner (2.63%) runs 43% higher than at lunch (1.84%). The empty table is disproportionately the one you could have sold twice on a Saturday night.

No-show rates by country

Twelve countries meet the publication threshold of at least 30 restaurants and 1,000 reservations. Together they cover 86% of the dataset.

Country	No-show rate	Reservations	Restaurants
Switzerland	0.49%	71,230	64
Germany	1.30%	58,523	51
Denmark	1.31%	28,606	30
Australia	1.41%	251,804	213
United Kingdom	1.48%	2,033,392	1,089
Sweden	1.75%	26,984	35
Spain	2.18%	104,047	65
Georgia	2.95%	54,131	46
Canada	2.98%	78,759	60
United States	3.17%	309,604	316
Italy	4.71%	118,488	45
France	5.88%	101,795	60

Country differences are descriptive. They can reflect restaurant mix, booking sources, how consistently staff record no-shows, and feature adoption, not only guest behavior.

Large parties are safer than you think

The highest no-show rate belongs to the smallest tables, and the lowest to the largest:

Party size	No-show rate	Avg covers per no-show	Expected lost covers per booking	Reservations	Restaurants
1 to 2 guests	2.63%	1.96	0.051	1,860,642	2,399
3 to 4 guests	2.13%	3.54	0.076	1,150,708	2,379
5 to 6 guests	2.28%	5.50	0.125	414,338	2,331
7 to 9 guests	1.73%	7.75	0.134	187,425	2,231
10+ guests	1.15%	13.44	0.154	155,583	2,122

The operator instinct to fear the big table is half right, though. Large parties no-show less often, but the ones that do take far more with them: a no-show in the 10+ bracket averages 13.44 covers, against 1.96 for a table for two. Multiply each rate by that size and the ranking inverts. Expected lost covers per booking rises steadily with party size, and a 10+ booking carries three times the exposure of a table for two (0.154 covers against 0.051). The rational policy is not refusing large bookings, it is protecting them.

Note that the 10+ bracket is open-ended, so its average is well above ten. Reading the last column off the rate and a literal party of ten understates the exposure by about a quarter.

Booking further ahead does not mean flakier guests

Received wisdom says early bookings are risky. The data says the opposite:

Booked ahead	No-show rate	Reservations	Restaurants
Same day	2.09%	1,458,490	2,353
1 day	2.40%	602,325	2,364
2 to 6 days	2.67%	907,550	2,370
7 to 13 days	2.57%	343,507	2,306
14 to 29 days	2.34%	270,985	2,243
30+ days	1.96%	183,329	2,021

Guests who plan a month out show up more reliably than guests who booked earlier the same week. The riskiest window is 2 to 6 days ahead. Lead time is measured in restaurant-local calendar days between booking creation and the service date.

Day, service, and season

• Monday is the worst day (2.69%), Wednesday the best (2.09%). All seven days published, 1,698 to 2,322 restaurants per cell.
• Dinner (2.63%) runs worse than lunch (1.84%), defined as service times from 17:00 and from 05:00 respectively.
• January is the worst month (2.62%), August the best (2.14%). New-year resolutions apparently do not extend to honoring reservations.

The hidden no-show: late cancellations

Of 566,049 cancellations in the dataset, 317,541, or 56%, were cancelled within 24 hours of the booking or after it had started. 40,360 were cancelled after the booking’s start time, which for planning purposes is a no-show with paperwork.

Cancelled	Share of cancellations
After the booking started	7.1%
0 to 2 hours before	17.4%
2 to 6 hours before	15.0%
6 to 24 hours before	16.6%
24 to 48 hours before	11.6%
2 to 6 days before	17.4%
7+ days before	14.6%

Shares are computed on all 566,049 cancellations; 1,642 records (0.3%) have a missing or invalid timestamp and are not shown as a row.

If your no-show policy ignores late cancellations, it addresses less than half the empty-table problem. A waitlist that automatically refills released tables recovers part of it.

How prevention features relate to no-shows

Everything in this section is an observed association, not a measured effect. Nobody was assigned to a group at random, so these numbers describe how bookings and restaurants that used a feature differed from those that did not. They do not establish that the feature caused the difference.

SMS reminders. Reservations with a delivered SMS reminder showed a 16% lower recorded no-show rate than reservations at the same restaurants without one (risk ratio 0.836, based on 46,884 exposed reservations at 257 restaurants that had reservations both with and without a delivered reminder). The comparison is between individual reservations inside the same restaurant, so it is an association at booking level, not a measurement of what happens when a restaurant switches reminders on. SMS reminders in Resos.

Deposits and no-show protection. Reservations requiring no-show protection recorded a 0.80% no-show rate (1,907 of 238,539 reservations at 327 restaurants) against 2.47% for reservations without a payment requirement (85,589 of 3,461,164 at 2,409 restaurants). We publish these as raw descriptive rates only: restaurants that require deposits differ from those that do not, and services protected by deposits differ from those left open, so this comparison cannot separate the feature’s effect from selection. We have since run a matched study that follows restaurants across the date they switched a payment feature on, comparing them with similar restaurants that did not. Every estimate it produced carries an uncertainty interval wide enough to include no change in either direction, so none of them is published here as a feature effect. We will publish one when a result survives replication on a held-out period, and not before. Online payments in Resos.

Confirmation prompts ask the guest to re-confirm before service. After accounting for restaurant differences they showed a near-neutral association in this dataset (risk ratio 0.983, based on 48,321 exposed and 688,234 control reservations at 459 restaurants).

Method and limitations

Population. Currently active Resos booking locations (2,417 with eligible outcomes), 1 August 2025 to 31 July 2026. One customer means one location with its own subscription.

Definition. A no-show is a staff- or integration-recorded outcome (no_show status), never an automatic timeout. The denominator is recorded no-shows plus attended reservations (approved, arrived, seated, and left statuses), counted only once the booking ended at least 72 hours before extraction. Walk-ins are excluded: a walk-in cannot fail to arrive for an advance reservation. Including them would read 2.06%.

What this rate is. A recorded rate, booking-weighted. 72.9% of eligible locations recorded at least one no-show; locations recording zero may be genuine zeros or non-users of the status. Among only the 1,761 locations that actively record no-shows, the rate is 2.49%. High-volume restaurants influence the headline more than small ones. The United Kingdom contributes 54% of reservations, so the global average leans toward UK behavior.

What this rate is not. Not a survey, not an estimate, and not directly comparable with survey-based figures about diners admitting to no-showing, which measure self-reported guest behavior rather than recorded outcomes at the reservation level.

Derived columns. “Avg covers per no-show” is the mean booked party size of the reservations that were recorded as no-shows in that bracket, not of all bookings in it. “Expected lost covers per booking” is the no-show rate multiplied by that average, so it answers “how many covers does an average booking of this size cost me” rather than “how often does it fail”. Both come from the same extraction as the rates.

Sample caveat. Resos customers are not a random sample of all restaurants, and every cell mixes restaurant type, booking source, and marking discipline. Cells with fewer than 30 restaurants or 1,000 reservations are suppressed, and their denominators are retained in the underlying extraction.

How to cite this data

Cite as “Resos No-Show Index 2026” and link to this page. Short form for articles: “According to Resos data covering 3.77 million reservations across 2,417 restaurants, the average recorded no-show rate is 2.33%.” You are welcome to quote the figures on this page with attribution. For press inquiries and custom cuts, see the press kit.

Frequently asked questions

What is the average no-show rate for restaurants?

According to Resos data covering 3.77 million reservations across 2,417 restaurants (August 2025 to July 2026), the average recorded no-show rate is 2.33% of advance reservations. Rates vary widely by country, from 0.49% in Switzerland to 5.88% in France.

Why is this lower than the no-show rates usually quoted in the press?

Most widely quoted figures come from consumer surveys asking diners whether they have ever skipped a booking, which measures lifetime incidence per person rather than a per-reservation rate. Platform operators have reported higher recorded rates, but those figures typically bundle late cancellations together with no-shows and count marketplace discovery bookings, which behave differently from a restaurant’s own direct guests. This index counts only the recorded no-show status on advance reservations, and reports late cancellations separately (56% of cancellations arrive within 24 hours of service). It is also a recorded rate: staff must mark the no-show, so the true rate is somewhat higher than the recorded one.

Which reservations are most likely to no-show?

In this dataset: parties of one or two (2.63%), bookings made 2 to 6 days ahead (2.67%), dinner services (2.63%), Mondays (2.69%), and January bookings (2.62%). Large parties and far-ahead bookings are more reliable than their reputation.

How do reminders relate to no-shows?

Reservations with a delivered SMS reminder showed a 16% lower recorded no-show rate than reservations at the same restaurants without one (46,884 exposed reservations at 257 restaurants that had reservations both with and without a delivered reminder). That is an observed association in operational data, not a controlled experiment, so it does not show that the reminder caused the difference.

How often is the index updated?

The underlying extraction is re-run quarterly, and a new edition of this page is published yearly with the same address, so citations and links stay valid.