Restaurant no-shows: what the UK data actually says
Restaurant no-shows are the most complained-about problem in British hospitality and one of the least measured. Only 19% of UK operators track their no-show rate at all (Propeller × KAM, Hospitality Bookings Benchmark, July 2026), which means most of the industry is furious about a number it has never calculated.
Here are the figures that hold up, the two everyone quotes that do not, and the connection between no-shows and your telephone that almost nobody makes.
How common are restaurant no-shows in the UK?
Around one booking in seven, and rising. Zonal and CGA by NIQ, surveying more
than 5,000 British adults, found the UK no-show rate climbing through 2024 — and
London running far worse than anywhere else, generating 28% of all UK no-shows
from 14% of the population, up from 24% in mid-2021.
Operators themselves report worse. In the Propeller × KAM benchmark published this month, 73% of operators said up to 1 in 10 of their bookings fail to show, with reported annual losses ranging from £8,000 to £400,000.
Figure | Source | Sample | Date |
No-show rate rose from 12% to 14% across 2024 | Zonal / CGA by NIQ | 5,000 hospitality consumers | 5 Nov 2024 |
London accounts for 28% of all UK no-shows, up from 24% in mid-2021 | Zonal / CGA by NIQ | 5,000+ British adults | Zonal / CGA |
73% of operators see up to 1 in 10 no-show; only a third can estimate the cost | Propeller × KAM | Not published | Jul 2026 |
27% of no-showers tried to cancel and could not reach the venue | CGA Consumer Pulse | Not disclosed | 4 May 2021 |
The London figure is the one to sit with. The capital has about 14% of the UK population and generates 28% of its no-shows. Londoners are roughly twice as likely as anyone else to book a table and not turn up.
Two numbers we will not repeat
You will see "£17.59bn a year in lost sales" everywhere. It originates in 2021 research with no published methodology and has not moved in three years of inflation.
You will also see "£89 per no-show" — Barclaycard, fieldwork May 2023, now three years old.
Both are quoted constantly in this industry. Neither survives a question about how it was worked out, and an operator who checks one will stop trusting everything else on the page.
What causes restaurant no-shows?
Mostly plans changing, and cancelling being harder than not bothering. The Propeller × KAM data gives the reasons diners offer: plans changed 59%, forgot 26%, someone else cancelled 21%.
But the finding that ought to change how you think about it is older, and it points somewhere unexpected. CGA's Consumer Pulse research found that 27% of people who failed to show had tried to cancel and could not reach the venue (published 4 May 2021; sample size not stated in the coverage, so treat it as indicative).
More than a quarter of your no-shows may have tried to do the right thing and been unable to.
What do restaurant no-shows have to do with your phone?
More than most operators think, and the July 2026 UK data makes the link directly. From the Propeller × KAM benchmark:
22% of consumers rang the restaurant to book and no one answered it.
22% abandoned a booking because they could not reach a real person.
43% of consumers have considered making a booking and changed their mind.
27% of bookings cannot be traced to any source at all.
So should a restaurant let AI answer the phone?
Yes — provided it never stands between a guest and a person. At one in the afternoon mid-service, or at eleven at night, the realistic alternative to an AI answering is not a human. It is a ringing phone or voicemail. That is where it earns its place. And the moment a guest asks for a person, it should hand them straight over, no loops and no insisting. That is how ours is built.
What works
Announce it is automated in the first breath. People forgive a robot for being a robot. They do not forgive it for pretending, and the complaints come from guests who work it out halfway through a sentence.
Use it for the calls that are genuinely routine — opening hours, parking, whether you do a set menu, a straightforward table for two.
Hand anything unusual to a human. Large groups, allergies, press, suppliers, and anyone who sounds annoyed.
Make cancelling effortless. Given that 27% of no-showers tried to cancel and could not, an AI receptionist that always answers is a no-show prevention tool, not just a booking tool.
What goes wrong
Accents and background noise. Guests who realise halfway through. And any system that tries to handle the awkward edge case instead of passing it on — which is a problem, because the awkward edge case is very often why someone picked up the phone rather than booking online.
If you take one design decision from this, make it the read-back: the system should repeat the booking details to confirm, and save the actual recording rather than only its own summary of it.
Is UK hospitality actually using AI yet?
Barely, and less than the headlines suggest. 20.6% of UK accommodation and food service businesses use AI — the fourth-lowest of thirteen industries measured (ONS, Artificial intelligence in UK businesses, published 20 July 2026, June 2026 data).
Adoption is also shallow. Of those using it, most run exactly one tool, and only 12.5% pay for software at all — the lowest of any industry in the release. UK hospitality's AI is overwhelmingly free chatbots, not purchased systems.
On bookings specifically, 27% of operators can use AI to take telephone bookings, and 35% have no plans to use AI at all (Propeller × KAM, July 2026).
Which is to say: if you are weighing this up, you are early. Not late.
Questions restaurants ask us
What is a normal restaurant no-show rate in London?
The best UK figure is around 7% of bookings (ResDiary, January 2025), with most operators reporting up to 10%. London runs materially worse than the rest of the country — 28% of all UK no-shows come from the capital despite it holding 14% of the population.
Will an AI receptionist annoy my guests?
Some of them, yes. 12% of UK consumers said they would abandon a booking rather than talk to a bot. The evidence suggests the resentment is aimed at machines pretending to be people, not at automation itself. Say what it is, keep it to routine calls, and hand the rest to a human.
Do deposits stop restaurant no-shows?
They help, and guests broadly accept them: 77% agree deposits are fair at busy times and the most-favoured amount is £6–£10 a head (Propeller × KAM, July 2026). But 31% said easier cancellation would stop them no-showing, which costs you nothing at all.
Why does nobody know the phone-versus-online booking split?
Because it has never been published for the UK. The only dedicated UK bookings study, released this month, has a section on where bookings come from and does not break it out. If a supplier quotes you a figure, ask them where it came from.
If your phone is the gap, we can close it. GPTable's AI receptionist for restaurants answers every call, books straight into the system you already use, and hands anything unusual to you. It also makes cancelling easy — which, on the evidence above, is a no-show problem as much as a booking one.
Sources: Propeller × KAM Hospitality Bookings Benchmark, fieldwork June–July 2026, sample sizes not published. Zonal / CGA by NIQ, GO Technology: the truth behind no-shows, 5 November 2024. Zonal / CGA by NIQ on London no-shows. CGA Consumer Pulse via Restaurant, 4 May 2021. ONS, Artificial intelligence in UK businesses, 20 July 2026.
Comments