How to design a restaurant's shift catalogue

Almost no schedule problem is fixed in the schedule. It is fixed earlier, in the shifts you build it from. If the blocks are badly cut, no combination saves them: people are missing right at the peak, there are too many in the mid-afternoon, and overtime appears on its own at the end of the month.

What a shift catalogue is

It is the restaurant's short, stable list of shifts, each with:

• A code the team reads at a glance: A, B, C, or whatever names make sense for your restaurant.
• Its hours, from start to finish.
• Its paid hours, which are the hours minus the meal break.
• The days it can be used, because a Thursday-to-Saturday close is not the same as a Monday-to-Wednesday one if the restaurant closes at a different time.

The month's schedule is built only from those shifts. Nobody invents a new set of hours on a Tuesday: you pick from the ones that exist. That keeps the schedule readable, the hours predictable and changes simple ("move Juan from B to C").

An example

A restaurant whose kitchen closes at 23:30 and whose dining room closes at 01:30, with a lunch peak and a dinner peak:

Seven rules for cutting good shifts

1. Arrive before the peak. A shift that starts at 13:00 to cover a lunch that starts at 13:00 doesn't cover lunch: the person arrives, changes and catches up when the dining room is already full. A 30-minute margin before the peak is a reasonable starting point; some restaurants use 15 and others an hour.

2. Never start or end in the middle of a peak. A shift that ends at 20:00 during a 19:00–22:00 dinner leaves the restaurant short exactly when it is needed most, and makes "how many people are at dinner" depend on when you look.

3. Prefer long shifts that span both peaks. A ten-hour shift that covers lunch and dinner replaces two short shifts. For a full-time contract, whose hours are already paid, it is the difference between filling the contract and paying for hours nobody works. The legal limit is ten ordinary hours a day: see the working day.

4. Open and close with just enough. Only the minimum needed has to come in early to open, and only the minimum has to stay to close. The rest of the dinner team can leave after the peak, before closing.

5. One closing shift per closing time. If the restaurant closes at 00:30 Monday to Wednesday and at 01:30 Thursday to Saturday, those are two different closing shifts. See the closing shift.

6. Few shifts, stable ones. A catalogue of eight to twelve shifts is usually enough for a restaurant with two services and weekend variants. Every new shift added "just this once" makes the schedule harder to read and to comply with.

7. Change the catalogue when the restaurant changes, not when the week does. New service hours, a closed day or a big change in staff justify revisiting the catalogue. Someone's holiday or a week with more bookings is solved in the schedule, not the catalogue.

History is not rewritten

When a shift's hours change, the right move is to create a new version and leave the old one in the weeks already worked or published. If the original shift is edited, past weeks start showing hours that were never worked, and the record stops matching what happened.

How Batuta handles it

In Batuta the catalogue is a separate piece, with codes, hours and paid hours with the meal break deducted. If you don't yet know which shifts you should have, Batuta can design them: it takes your coverage, your staff and your contracts, and proposes shifts that cover the service, avoid overtime and use the fewest people. It then shows exactly what changes against your current catalogue, and you decide whether to adopt it. Shifts are versioned, so what was published keeps showing as published.