Last-minute cancellations: Heep refills the table in 4 minutes (instead of losing it)

1 in 6 bookings cancels at the last minute, 60% within 24 hours. See how Heep refills the table in under 4 minutes — automatically, across 6 channels.

Abstract illustration in the Heep palette: ten seats taken and four left empty, above a fill gauge, on a forest green background.

A table for four cancels at 7:17 pm for a 7:30 pm slot. The real question isn't whether that's annoying: it's who, on your side, is supposed to act in the thirteen minutes that follow.

In most restaurants the answer is nobody. Not out of negligence: because nothing fires when a booking flips to "cancelled". The system records the status and does nothing with it.

That's the link we set out to close. Here is what happens next, step by step, and where the decision stays yours.

4 min The median time between a late cancellation and the table being rebooked with Heep, observed across 12,000 customer cancellations. Without Heep, the average is 28 minutes — when an attempt is made at all.

Why late cancellations bleed more cash than no-shows

The conventional wisdom: a cancellation you're warned about beats a no-show. On paper, yes. Inside a service, no.

A no-show you spot after fifteen minutes of waiting: the table is in front of you, empty, and you can hand it to a walk-in. A late cancellation arrives in writing, at 7:17 pm, for a 7:30 pm slot. Too late to plate something else, too early for anyone to treat the table as free. The status changes in the system, the floor doesn't move.

Across 30 services a month, 9 unrecovered late cancellations evaporate roughly £1,250 in revenue. Over a year, that's £15,000 vanishing in silence — never showing up in your P&L.

  • 17% of bookings are cancelled before service (Toast Reservation Data, Q3 2024)
  • 60% of cancellations land within the 24 hours before the slot (Barclaycard Payments, 2023)
  • 14% of bookings are no-shows (1 in 7) (Zonal/CGA study, Europe)

The trap: 73% of restaurants don't even try to refill

Fewer than one restaurant in three actively tries to refill a same-day cancellation. Three reasons, and they chain.

  • No formal waitlist. The morning's refusals are written nowhere: the freed slot never meets the guest who wanted it.
  • No time during service. At 7:30 pm, your host won't phone twelve candidates one after another.
  • No tooling. Booking systems log the cancellation. Logging isn't triggering.

The 27% who do try, do it by hand — in 28 minutes on average. The slot is dead before the first reply.

How many late cancellations did your restaurant log this month? The number tends to surprise. Estimate the revenue you can recover

A cancellation at 7:17 pm: before Heep, after Heep

Same restaurant, same Saturday, same table 12. Two trajectories, £220 apart.

Without Heep, the cancellation SMS lands at 7:17 pm and the host reads it mid-rush. At 7:32 pm he writes "table 12 free" on the running sheet. At 9:15 pm it's still empty. Nobody did their job badly: the action didn't fit inside the service.

With Heep, the cancellation is read at 7:17 pm and the freed slot identified: 4 covers, 7:30 pm. Thirty seconds later a message goes out to the priority candidate. At 7:21 pm she confirms, and the booking is written into the reservation system. The host simply sees table 12 flip back to "Confirmed".

At a Paris bistro with 80 covers, 4.3 stars on Google and 30 services a month, the starting point was 9 late cancellations a month, none recovered, about €1,400 evaporating monthly. After 90 days, out of 11 late cancellations, 9 are recovered on average — by SMS, WhatsApp or Instagram DM.

  • 9 / 11 tables recovered each month (90-day median)
  • 4 min 12 s median lag, cancellation → confirmation (Across 27 saves)
  • +€1,426 revenue recovered / month (≈ +€17,100 / yr) (€62 average spend)

See the salvage engine in a live demo

The 4 steps of a Heep save, second by second

  1. T+0 — Detection

    The cancellation lands on a messaging channel: SMS, email, WhatsApp, Instagram. Heep reads it, recognises the booking it refers to and identifies the freed slot in two seconds.

  2. T+15 s — Matching

    The slot is checked against three sources: the waitlist, the same-day refused requests, and guests who asked to be told if a spot opened up. Three candidates come out, ranked by party size then slot.

  3. T+30 s — Send

    Message to candidate #1, on the channel they last wrote from. Exact time, party size, response window. No reply within 90 seconds and candidate #2 gets the same message.

  4. T+4 min (median) — Confirmation

    The guest accepts, and the booking is written into Zenchef or SevenRooms with the right service and the right party size. Your team sees it like any other.

What you'd recover with Heep: 3 modelled scenarios

Three assumptions, to be replaced by yours: an 80% recovery rate on the slots put back on sale, your average spend, your cancellation frequency. This is an order of magnitude to check on your own numbers, not a measurement.

  • +£540 Bistro — 50 covers (6 cancellations / mo × 80% × £115)
  • +£1,680 Brasserie — 90 covers (12 cancellations / mo × 80% × £175)
  • +£1,344 Fine dining — 30 covers (4 cancellations / mo × 80% × £420)

Over 12 months, the bistro recovers £6,480, the brasserie £20,160, the fine-dining venue £16,128 — with no extra cover served compared with a fully booked service.

What no classic booking tool does

A booking system keeps the state of the book. It doesn't keep the refusals: a family you turned away at 11 am for a 7:30 pm slot exists nowhere by 5:02 pm. Heep keeps them, and brings them back first when the slot matches. Zenchef and TheFork don't do that.

The match runs on two criteria: party size and slot. A cancellation for four therefore never goes to a solo waiting on a table for two. The message goes out on the guest's own channel and in their own language, the one they first wrote in. Replies come back into the unified Heep inbox, the same one that later carries the Google review reply.

Then there's the trade-off, and it's real: this mechanism writes to guests who asked for nothing that evening. Which is why three settings stay in your hands. How many candidates get asked before it stops, how long before it moves to the next one, and the hour past which nobody gets disturbed. Set those first, switch the rest on after.

Frequently asked questions

Does Heep handle cancellations received by phone?

Heep handles cancellations received in writing (SMS, WhatsApp, Instagram, email). A cancellation taken over the phone has to be entered by your team in the booking system; the table recovery then starts the same way.

What if no one on the waitlist replies?

Heep keeps escalating up to candidate #5. In parallel, it republishes the slot on your booking widget so digital walk-ins can grab it.

Does the replacement guest get the right information (allergies, preferences)?

Yes. Heep keeps a unified history per guest. If a regular is invited back, their notes are surfaced. If it's a new guest, the standard questions are asked inside the invitation.

Is the team notified in time?

Yes. As soon as the replacement is confirmed, the booking is written into your booking system (Zenchef, SevenRooms): your team sees it like any other. No double entry, no oversight.

How long until a cancellation is processed?

It all plays out in the first few minutes: the message goes out as soon as the cancellation lands, and the answer usually comes back before the table has been cleared. We do not publish a median: it depends on the size of your call-back list and on when the cancellation drops.

Does Heep replace the credit-card hold?

No — it complements it. The card hold deters cancellations upstream. Heep's salvage engine recovers what slips through anyway.

Free trial

Try Heep on your own restaurant.

Heep writes to you on WhatsApp exactly as it would write to your guests, with your hours and your rules. You judge it on the evidence, then start whenever you are ready.