Managing walk-ins vs bookings: the yield strategy for restaurants and pubs

Count seat-hours before you count tablesLink

Revenue per available seat-hour is the only number that matters.

Start with the only number that matters: revenue per available seat-hour. A four-top occupied for two hours is not the same asset as a four-top occupied for 75 minutes. A table that appears “full” may be producing less revenue than a faster-turning table elsewhere on the floor.

Map your usable capacity by seats, table sizes, expected party sizes, average dining duration, kitchen production capacity, bar and front-of-house capacity, and the number of turns available in each service. Do not use theoretical capacity — if your kitchen can only handle 18 covers in a 30-minute window, that governs your usable capacity, not the chairs.

Use this baseline: available seat-hours = usable seats × trading hours. Then measure how much of that inventory produces revenue by tracking empty table minutes during peak periods, covers by booking and walk-in source, spend per head by source, average table duration, no-shows and late cancellations, walk-ins turned away, and walk-ins who accept or reject the quoted wait.

If you are not recording those numbers, you are not managing yield. You are guessing.

Set the booking mix by daypartLink

Tuesday lunch and Saturday dinner are different products.

Never use one reservation policy for every shift. Use a starting allocation, then adjust it with your own data: weekday lunch at 30–40% reservations and 50–60% walk-ins; weekday dinner at 50–60% reservations; Friday and Saturday dinner at 60–70% reservations; pub-led or bar-led service keeping a larger walk-in allocation; and Sunday roasts or major events with increased reservation capacity and deposits for larger parties.

These are starting points, not commandments. The mistake is booking the whole room because the system allows it. Do not book to 100% of theoretical capacity — hold back inventory for walk-ins, same-day demand and the overruns that happen in real hospitality.

Protect bar seats, high-top tables, smaller two-tops, a flexible section for parties of two to six, and tables that can be released later in the service. Keep the rules in the booking system rather than relying on one manager's memory. A clear, mobile booking path also matters — a customer should not have to hunt for the booking button or make four clicks to find a phone number.

• If your booking flow needs work, see HMRA's restaurant marketing services
• or book a free audit

Protect the walk-in bufferLink

A walk-in buffer is a deliberate sales channel, not spare capacity.

The exact percentage depends on your venue, but 20–40% of usable capacity is a sensible starting range for many independent pubs and casual restaurants. Adjust it by service, season and local behaviour.

Make the allocation physical. Mark the floor plan as reservation-only, walk-in-only, flexible, or held for a later booking. Then add release rules — for example, a table reserved for 7:30pm may be used for a walk-in before 6pm; a table held for a booking is released 15 minutes after the booking time if the guest has not arrived and has not called; a walk-in seated at a future-booked table must be told the return time before ordering; and the host can override the rule only with manager approval.

This stops the two most common failures: empty reserved tables while walk-ins leave, and walk-ins seated without enough time to finish comfortably before the next booking.

Pace arrivals against the narrowest constraintLink

Your booking system sees tables. Your operation has a kitchen, bar and service team.

Set arrival limits in 15- or 30-minute windows based on the narrowest constraint: table availability, kitchen ticket capacity, bar production, number of servers, expected party size and existing table overruns.

If the kitchen is already buried, stop seating just because empty tables are visible. An empty table with a 45-minute ticket time is not available capacity — it is a guest-experience trap. Stagger bookings instead of accepting a wall of arrivals at 7pm: for example 12 covers at 6:30pm, 8 at 6:45pm, 12 at 7pm, 8 at 7:15pm and 10 at 7:30pm.

The exact numbers depend on your kitchen. The principle does not: pace the room to what the team can serve properly. Maximum occupancy is not maximum revenue if the result is slow food, abandoned drinks and poor reviews.

Set a return time before seating, and manage the bar as a waiting assetLink

Honesty about table turns protects both the guest and the team.

A table booked for 8:15pm is not available for an unrestricted two-hour meal at 6:45pm. Tell the guest before seating: “We can seat you now, but we need the table back by 8pm.” That is honest, not rude. Use dynamic hold times based on party size, kitchen pressure, table location, the next reservation and whether the guest is eating or drinking. Do not promise a 60-minute turn when your data says the average is 85 minutes.

The bar should not become a punishment zone. Managed properly, it is a controlled waiting area that protects guest experience and generates additional spend. Give waiting guests a realistic quoted wait, somewhere comfortable to sit or stand, a digital waitlist, a text notification, a drinks and small-plates offer that does not destroy margin, and a visible host who owns the process.

Do not quote “about 20 minutes” because it sounds better — a false promise creates walkaways and complaints. Set a maximum waiting-area occupancy and measure waitlist numbers, acceptance rate, average wait, drinks revenue while waiting, and the percentage leaving before notification.

Release held tables before demand expires, and match occupancy to guest experienceLink

Full is not always good.

A reservation is not revenue until the guest arrives. Confirm bookings 24–48 hours ahead, use card holds or deposits for high-demand periods, contact late parties quickly, and release no-show tables after 15 minutes unless the guest has made contact. Do not keep a table empty for 40 minutes because the system says “reserved” — that is unpaid storage, not service. Deposits make sense for large groups, weekend evenings, Christmas, Sunday roasts, special menus and events, and guests with a history of no-shows. Apply the policy consistently.

An overloaded dining room produces longer ticket times, more mistakes, slower payments, staff stress, lower drinks and dessert spend, bad reviews and more service recovery. A room at 92% effective occupancy with clean service can outperform a room at 100% physical occupancy that collapses under pressure.

Use a red-line rule: when the kitchen, bar or service team hits its limit, reduce arrivals, quote longer waits, pause walk-ins and slow online bookings. That is margin protection — a poor review can cost more than the few extra covers you forced through the door.

Track the numbers every weekLink

Judge the strategy by the data, not the booking diary alone.

Review these metrics every week by day and service: reservation fill rate, walk-in covers, walk-in conversion rate, no-show rate, cancellation rate, average spend per head, average table duration, empty table minutes, walkaway rate, revenue per available seat-hour, labour cost per cover, kitchen ticket times and drinks spend while waiting.

Then change one operating rule at a time. If reserved tables sit empty while walk-ins leave, tighten release times or reduce reservation inventory. If you turn away bookings but finish with empty tables, fix table mapping and arrival pacing. If walk-ins spend less, improve drinks, bar snacks and dessert conversion rather than removing walk-ins automatically — they may still deliver better seat utilisation. If Saturday is full but Tuesday is dead, put demand generation against the empty shift rather than spending more on Saturday.

• HMRA's restaurant revenue management service
• covers demand forecasting, table-plan reviews and no-show controls — see also our guide to cutting no-shows
• and booking page conversion leak audit