A covers forecast your kitchen will believe

A covers forecast is an estimate of how many guests the venue will serve at each service. For the kitchen, it is the number that decides how much to prepare, how much to order and how the stations should be set up. When the kitchen believes the forecast, preparation is calm, waste is low and the kitchen is ready for the rush. When it does not, cooks prepare from instinct, usually too much, sometimes too little, and the forecast becomes a number written in an office that nobody downstairs reads.

Kitchens stop believing forecasts for understandable reasons. The forecast was too low on a busy Saturday and they ran out of the best dish by half past eight. Or it was too high on a wet Tuesday and they threw away trays of prepared vegetables. After a few of those, the head chef quietly adds a margin of their own, or ignores the number altogether.

This guide sets out how to build a covers forecast the kitchen will believe. It covers bookings, expected walk-ins, no-shows and cancellations, waitlist data, the sensible pad that protects against the busy night without wasting food on the quiet one, how to turn the forecast into prep from forecast, and how to check it every week so trust is earned rather than demanded.

The general idea of demand forecasting, estimating how much customers will want before they ask for it, applies to every kitchen. A sales forecast, in money, helps the owner and the accountant. The kitchen needs covers, by service, because covers decide how many plates go out. A sales figure of a certain amount on Saturday tells the chef little; a forecast of 95 covers at dinner, with a typical mix of starters and mains, tells them a lot.

The kitchen also needs it early enough to act on. Orders often go to suppliers days ahead, and preparation starts in the morning. A forecast that arrives at four o'clock on the day is too late.

Our pillar guide to sales forecasting for next week in twenty minutes covers the wider method. This guide focuses on the kitchen's number.

For venues that take bookings, the starting point is the covers already booked for each service. The diary is a direct signal of demand, and it is the part of the forecast the kitchen is most likely to trust, because it is made of real guests with real names.

Look at bookings by service, not just by day. Saturday lunch and Saturday dinner are different services with different preparation. Look at party sizes too: a service with several large parties may need different preparation from one with many couples.

Remember that the diary keeps changing. A Saturday that has 60 covers booked on Wednesday may have 80 by Saturday morning. That is why the forecast must allow for bookings still to come.

Most services keep filling after the forecast is made. Guests book a day or two ahead, or on the same day. To forecast the final number, you need to know how bookings usually build up: how much of a typical service's final covers is booked by the time you make the forecast.

Look at the last few weeks. If a typical Saturday dinner usually has about 70 per cent of its final booked covers by Thursday, and this Saturday has 63 booked on Thursday, you can expect around 90 booked covers by Saturday evening. This booking pace differs by day and service, and it changes with the season.

Many booking systems can show how much notice guests give, which is exactly what this calculation needs.

Walk-ins are guests who arrive without a booking. Some venues see very few; others, especially those with passing trade, bars or terraces, see many. Add an allowance for walk-ins to each service, based on the average number of walk-ins for that service over recent weeks.

Walk-ins vary with the weather, local events and the time of year. A terrace venue on a sunny Saturday may see far more walk-ins than on a cold one. Adjust the allowance for what you expect.

If you do not record walk-ins separately from bookings, start. Even a simple count at the host stand each service makes the forecast much better.

Not every booking arrives. Some guests cancel, some change their plans, and some simply do not turn up. Subtract an allowance for no-shows and late cancellations from each service, based on recent rates for that service.

No-show rates often differ by service, party size and how far ahead the booking was made. A large party booked weeks ahead may behave differently from a couple booked that morning. Our guide to your no-show rate covers how to measure it fairly.

A cancelled table may also be refilled from the waitlist, so the allowance should reflect the net effect, not just the cancellations.

Waitlist data tells you how much demand the venue turned away or delayed. On busy nights, a long waitlist means the venue could have served more guests if it had more room, or more guests if tables had turned faster. On quieter nights, an empty waitlist means the forecast should rely on bookings and walk-ins alone.

For the kitchen, the waitlist matters because guests seated from it are extra covers on top of bookings. A venue that regularly seats several waitlist parties on Saturdays should include them in the Saturday forecast.

Our guide to waitlist management covers how to run and record the waitlist.

The covers forecast for each service combines these pieces:

For example, for a Saturday dinner: 63 booked by Thursday, plus about 27 more expected, plus about 10 walk-ins and waitlist covers, minus about 5 no-shows, plus 3 for a local event, gives a forecast of about 98 covers. The figures are illustrative.

A pad is a small margin added to the forecast for preparation, to protect against running out on a busy night. Kitchens add pads anyway; the question is whether the pad is sensible and agreed, or large and hidden. A hidden pad on top of a forecast that already includes walk-ins leads to waste every quiet night.

Agree the pad openly. A modest margin, such as a few extra covers' worth of the most popular dishes, protects against an unexpectedly busy service without preparing for a crowd that will not come. The pad can be larger for dishes that are quick to prepare and keep well, and smaller for dishes that spoil quickly.

When the pad is agreed and the forecast is accurate, the kitchen has no reason to add its own.

Prep from forecast means turning the covers forecast into a preparation plan: how many portions of each component to prepare for each service. Multiply the forecast covers by the expected share of guests ordering each dish, from the recent menu mix, and add the agreed pad.

For example, if a forecast of 98 covers usually means about 25 orders of the lamb, the kitchen prepares for around 25 portions plus a small pad, rather than a round number from memory. Our guide to the prep list that ends the morning guesswork covers how to build the list.

Our guide to menu mix covers how to find each dish's share.

Send the kitchen two versions: a weekly forecast for ordering, a few days ahead, and a daily update for preparation, the evening before or first thing in the morning. The weekly forecast lets the chef place orders; the daily update reflects the latest bookings.

Make the forecast easy to read: one line per service, with covers, and a note of anything unusual, such as a large party, an event or an allergy-heavy table. A forecast buried in a spreadsheet the kitchen cannot open is a forecast the kitchen will not use.

The covers report, a simple record of actual covers by service, is how the forecast earns trust. Every week, compare the forecast with the actual covers for each service and note the difference. Share the comparison with the kitchen.

When the forecast is consistently close, the kitchen starts to rely on it. When it is wrong, the comparison shows where and why: too few walk-ins allowed on sunny weekends, too many no-shows on Tuesdays, a booking pace that changed with the season. Fix the method, and tell the kitchen what changed.

Our guide to the covers report, read on a Monday morning covers what the report should show.

Trust also comes from conversation. Ask the head chef which services they think the forecast gets wrong, and why. Cooks often notice things the numbers miss: a local office that books large lunches without warning, a dish that sells far more on Fridays, a regular group that always arrives on the first Thursday of the month.

Build what they tell you into the forecast. A forecast that reflects the kitchen's knowledge, as well as the diary, is one the kitchen is far more likely to believe.

Large parties and events need special handling. A party of twenty can change a service's preparation completely, especially if they order from a set menu. List large parties separately in the forecast, with their menu if known, so the kitchen can prepare for them specifically.

Confirm large parties a few days ahead, and update the forecast if numbers change. A party of twenty that becomes twelve, or thirty, changes what the kitchen needs.

The kitchen's busiest moments depend not only on how many guests come, but on when they arrive. A forecast of 98 covers spread evenly from six to ten o'clock is a very different service from 98 covers arriving between eight and half past. For the kitchen, the shape of the service matters as much as the total.

Add the expected arrival pattern to the forecast for busy services: how many covers are booked into each half hour, plus a typical pattern for walk-ins. That lets the chef set up stations, schedule cooks and time preparation for the real peak. Our guide to pacing covers covers how to spread bookings so the peak never overwhelms the kitchen.

Two numbers show whether the forecast and the pad are right: how much is wasted and how often dishes run out. Track both for each service. If quiet nights regularly end with trays of prepared food thrown away, the forecast or the pad is too high. If busy nights regularly see the best dishes sold out by half past eight, it is too low.

Share these numbers with the kitchen alongside the forecast comparison, every week, in the same simple format, so changes are easy to spot. They are the kitchen's own measure of whether the forecast works, and they make the conversation practical rather than abstract. Our guide to reducing food waste in a restaurant kitchen covers waste more broadly.

Lunch forecasts often behave differently from dinner. Fewer guests book ahead, more walk in, and the mix of guests depends heavily on the day of the week: offices nearby, shoppers, tourists, business lunches. A lunch forecast built on bookings alone will usually be far too low.

Rely more on the recent baseline and the walk-in allowance at lunch, and watch how weather and local events affect it. If the venue has a set lunch, its share of covers helps the kitchen plan preparation precisely.

Decide who owns the covers forecast. In many venues it is the manager, who knows the diary and the front of house; in others it is the head chef, who knows the kitchen's needs. Either works, as long as one person is responsible, it is made at the same time each week and both the floor and the kitchen see it.

The owner of the number should also own the weekly comparison with actual covers. A forecast that nobody checks never improves.

The pieces of the forecast change with the seasons. Booking pace shortens in summer, when guests decide at the last minute, and lengthens before the festive season, when they book weeks ahead. Walk-ins rise when the terrace opens and fall when it closes. No-show rates can change around holidays. A forecast built on last month's patterns can drift wrong as the season turns.

Review each piece at the start of every season: the booking pace, the walk-in allowance, the no-show rate and the menu mix. A short check every few months keeps the forecast in step with how guests actually behave at that time of year.

When the menu changes, the menu mix changes too, and preparation based on the old mix will be wrong. For the first weeks of a new menu, prepare a little more cautiously for new dishes, watch what sells, and update the shares as soon as you have a week or two of data.

Talk to the kitchen before the launch about expected sellers, and agree a slightly larger pad for the first few services.

Here is an illustrative week of covers forecasts and actuals, to show how trust builds. The figures are examples.

In servearound, the book shows what is already booked for each service, with party sizes, and the live waitlist shows each party waiting, its size and how long it has waited. Reservation analytics shows covers day by day and by meal period, how much notice guests give, bookings by source and party size, no-show and cancellation rates, and the share of booked covers that were seated, which are the pieces a covers forecast is built from.

The menu report shows how many of each dish sold, so covers can be turned into prep from forecast, and kitchen analytics shows how each station coped. See the reservations module and Analytics.

A weekly forecast a few days ahead for ordering, and a daily update the evening before for preparation.

Yes, from the average walk-ins for each service, adjusted for weather and events.

Small and agreed, larger for dishes that keep well and smaller for those that spoil.

Start counting them at the host stand each service. Even a rough count improves the forecast.

Compare it with actual covers every week, share the results and fix the method where it is wrong.

List them separately in the forecast, with their menu, and confirm numbers a few days ahead.

Yes, if the same kitchen cooks them. Count them by service alongside dining room covers, so the kitchen sees its whole workload.

Look for the reason, such as an unexpected large walk-in group or a sudden storm, note it, and adjust the method only if the reason is likely to recur. A single unusual night should not change the whole forecast, but a pattern should.

Yes. The weekly covers forecast, with the expected menu mix, tells the chef roughly how much of each ingredient the week needs, which makes orders more accurate and reduces both waste and shortages.

A covers forecast the kitchen will believe starts with bookings by service, adds bookings still to come from the usual booking pace, adds expected walk-ins and waitlist seatings from waitlist data, and subtracts no-shows and late cancellations. Adjust for weather and events, list large parties separately, and agree a small, open pad instead of a hidden one.

Turn the forecast into prep from forecast using the menu mix, send it early enough to act on, and keep a covers report that compares forecast with actual every week. Share the comparison with the kitchen, build their knowledge into the method, and let accuracy, not authority, earn the kitchen's trust in the number.