How Should Bookings and Functions Change What My Kitchen Orders and Preps?

If 140 covers are booked this Saturday against 90 last Saturday, last Saturday’s order is wrong before anyone writes it. Restaurant demand forecasting from bookings starts with a normal day’s sales, scales it to booked covers and walk-ins, and gives every confirmed function its own order and prep plan. Get it right and nobody gets sent to the supermarket at 4pm.

Most chefs already do this in their heads. The problem is that the bookings live in one system, the function sheet in an email, and the order in a group text with the rep. By Thursday the Saturday wedding is "on the list somewhere", the usual weekend order goes in from par, and the brisket for 80 arrives a day late or not at all.

Why isn’t sales history enough on its own?

Sales history tells you what a normal day looks like, but it cannot see what is already booked. Last Saturday’s sales are a good starting point for this Saturday only if this Saturday looks the same.

History misses:

• Booking swings. 140 booked covers this Saturday against 90 last Saturday is a different shift.
• Functions. A 60-guest set menu barely shows in an à la carte average, but it can double your use of one protein.
• Public holidays and events. A long weekend, a school holiday or a big game nearby moves walk-ins.
• Menu changes. A new special has no history at all.

Getting this wrong is costly in catering-heavy venues. End Food Waste Australia reports that in catering, 20% of the food served is never eaten. Some of that is portioning, but a lot comes from cooking and ordering to a guess rather than a confirmed number. If you still order mostly from par, our guide to par levels vs forecast ordering shows which lines to move first.

How do booked covers change restaurant demand forecasting?

Booked covers set the floor for the shift, and your usual walk-in rate fills in the rest. A simple covers forecast is booked covers plus expected walk-ins, with walk-ins taken from history for that day of the week.

A worked example

These numbers are made up. Say a bistro looks at its last six Saturdays and finds it averaged 70 booked covers and 30 walk-ins, so walk-ins added about 43% on top of bookings. This Saturday it has 95 booked.

• Expected walk-ins: 95 × 0.43 ≈ 41
• Forecast covers: 95 + 41 = 136
• Usual Saturday: 100 covers

So the forecast is 36% above a normal Saturday. If the bistro usually sells 22 lamb shoulders on a Saturday, a first estimate is 22 × 1.36 ≈ 30.

In Excel, put booked covers in B2, the historical walk-in ratio in B3, normal-day covers in B4 and normal-day portions of the dish in B5:

Two cautions. First, walk-ins do not always scale with bookings; on a fully booked night they drop because there are no tables, so cap the forecast at your seats and realistic turns. Second, no-shows matter. If you lose 5% of bookings on average, take that off first.

Forecasting like this is moving into software. In the Lightspeed APAC State of the Hospitality Industry Report 2026, as reported by Australian Bartender, 56% of venues said forecasting and optimisation tools save them more than an hour a day. That is a vendor survey, so treat it as a direction, not a guarantee.

How do you order and prep for a function?

Order and prep a function from its confirmed menu and guest numbers, on its own timeline, not from par. You have the menu, the guest count and the date, which makes a function the most forecastable part of the week. It is still often the part that goes wrong.

The steps:

1. Lock the final numbers. Agree a cut-off with the client, often 7 days out, for final guest numbers and dietaries.
2. Explode the menu. For each course, multiply the recipe per portion by guests. If you add an allowance for staff meals and spoilage, do it the same way every time.
3. Keep the function order separate. Function lines go on their own order, timed to land the day before prep starts, not on the normal par day.
4. Build a prep timeline back from service. Work back from when each course goes out.
5. Check overlaps with normal service. A Saturday wedding and a full restaurant compete for the same oven, combi and people.

An example prep timeline

Say a venue has a confirmed 80-guest wedding on Saturday, with braised beef cheek as the main. Working back from service, the week might run like this:

Write it per station so each section knows what is theirs. A forecast-based prep list built from what you will actually sell keeps the normal service from being squeezed, and how prep sheets conduct kitchen harmony covers writing sheets people follow.

Getting function waste under control pays. A Champions 12.3 review of 86 catering sites across six countries found the average site earned a 6:1 return on what it spent cutting food waste, and 80% of sites had recouped their investment within two years. Those are overseas sites, but caterers everywhere bin the same over-cooked trays.

What if bookings live in a different system?

If bookings live in a different system, pull a daily number out of it rather than trying to make the systems talk. Most booking platforms can export covers by day or publish a calendar feed, and function enquiries usually arrive by email.

Practical options, simplest first:

• Manual: each morning, write tomorrow’s and the weekend’s booked covers onto the order sheet.
• CSV: export bookings weekly and paste them into a tab that feeds your forecast.
• Calendar feed: subscribe to the booking system’s calendar so covers appear alongside your order days.
• Forwarded email: set a rule so confirmed function sheets land in one inbox the head chef checks before ordering.

Whichever you pick, make one person responsible for checking that the functions on the calendar match what is on order. Most function stuff-ups are communication failures, not cooking ones.

How Restaurant AI Manager helps with bookings and functions

Restaurant AI Manager, Open Pantry’s AI restaurant operations software, lets you add bookings and functions by hand, by CSV, by calendar feed or by forwarded email. (Open Pantry is an Australian ordering and operations platform for food and beverage suppliers and the restaurants they supply, founded in 2018.) They feed straight into the forecast, the orders and the prep.

Forecast orders combine sales, bookings and functions, subtract what is on hand and on order, round up to packs and are timed to each supplier’s delivery days. Nothing is sent until you approve it. Prep lists are built by station and rounded to batch sizes, and a confirmed function, such as an 80-guest wedding, gets its own prep timeline. See how it fits catering and functions, or try the free live demo with no sign-up.

In summary: ordering and prep from bookings and functions

Start with sales history for a normal day, scale it to booked covers plus your usual walk-ins, and plan every confirmed function on its own order and prep timeline. Restaurant demand forecasting from bookings means ordering to what is coming, not to what happened last week.