Guestit
English
ContactLog inIntroduction

Operations

9

min read

Hospitality staff planning: roster on hourly demand

Plan your roster on the work per hour instead of daily revenue. With three signals, a worked example and a four-Friday test.

Illustration: two staff members in aprons confer over a tablet at the bar of a restaurant that is still empty

Do this today

Take your roster for this coming Friday and write down, per hour, how many people are working and in which position: bar, floor, pass and kitchen.

Picture this: at 17:00 there are four people on the floor and three tables are occupied. At 19:00, six tables arrive within twenty minutes. There are still four people working, but suddenly everyone is waiting at the same time for a drink, an order or the bill.

At the end of an evening like that, the ratio between wage costs and revenue can look perfectly fine. Meanwhile, the roster had a problem you cannot see in the daily total: at one moment you had capacity to spare, at another you were short of hands.

Staff planning in hospitality therefore starts with a different question than "how many people do I need on Friday?". How much work hits your team at which hour, and who has the experience and the time to handle it? In this article you will read which three signals to watch, see in a worked example how the same hours can give you different staffing, and get a four-Friday test and one action for today. The point is to distribute the same hours better, not to use fewer hours.

Revenue tells you how much you sold, the order how much work it was

Two evenings can both bring in €4,000 in revenue and still call for very different staffing. That amount is made up; what matters is the difference. On one evening, large groups sit at their tables for a long time and order a set menu. On the other, you have lots of small tables, individual cocktails and guests who pay separately. For the bar and the floor staff, that can feel like two different services.

If you only look at revenue for your roster, you miss that difference. How to plan hours on the revenue you expect per part of the day is covered in Demand-based rostering: plan on expected sales. This article goes one step further: from revenue per part of the day to work per hour.

Three signals for your hospitality roster

Look beyond daily revenue for your roster. Start with three signals:

  • Reservations and walk-ins: how many guests arrive per quarter of an hour or half hour? And how many tables are still occupied at that point by guests who came earlier?

  • Orders per moment: when are the first drinks ordered, when do the main courses come in and when does the kitchen have to send out several tables at once? Look at the moment of ordering, not the moment of payment. A bill paid at 21:30 may have created most of its work between 19:00 and 20:00.

  • Work without a sales receipt: mise en place, washing up, setting up the terrace and closing also take time. Include those hours, even if there is little or no revenue at that moment.

If you have no overview of order times, you can keep a tally for one evening: for each half hour, make a mark for every order that goes to the bar or the kitchen.

Worked example: same hours, better staffing during the peak

This is a worked example with made-up numbers, not the roster of a real venue. Suppose a restaurant has two permanent floor staff on Friday who work from 16:30 to 22:30, and two extra staff members with a four-hour shift. In this example, the busiest period on previous Fridays falls between 18:30 and 20:30. Even so, the existing roster has the first extra staff member start as early as 16:30, while the second does not arrive until 19:30.

Existing roster

  • Two permanent staff: both 16:30 to 22:30, 2 times 6 hours is 12 hours

  • Extra staff member 1: 16:30 to 20:30, 4 hours

  • Extra staff member 2: 19:30 to 23:30, 4 hours

  • Total scheduled hours: 12 plus 4 plus 4 is 20 hours

Adjusted roster

  • Two permanent staff: both 16:30 to 22:30, 2 times 6 hours is 12 hours

  • Extra staff member 1: 17:30 to 21:30, 4 hours

  • Extra staff member 2: 18:30 to 22:30, 4 hours

  • Total scheduled hours: 12 plus 4 plus 4 is 20 hours

Staffing per hour

  • 16:30 to 17:30: existing 3, adjusted 2

  • 17:30 to 18:30: existing 3, adjusted 3

  • 18:30 to 19:30: existing 3, adjusted 4

  • 19:30 to 20:30: existing 4, adjusted 4

  • 20:30 to 21:30: existing 3, adjusted 4

  • 21:30 to 22:30: existing 3, adjusted 3

  • 22:30 to 23:30: existing 1, adjusted 0

  • Check: if you add up the staffing per hour, both rosters come to 20 hours again

In this made-up example, the restaurant gets an extra pair of hands in the first hour of the peak, from 18:30 to 19:30, and also in the hour after the peak, with the same number of scheduled hours. The trade-off is at the edges of the evening. Before 17:30 there is one person fewer, and after 22:30 there is no one left on the floor in the adjusted roster.

Whether that works depends on your venue. If you need the first extra staff member early for preparation, or someone has to be there until 23:30 to close, solve that in the roster first. The example also counts rostered hours: a six-hour shift includes a break, and it has not been deducted here. Schedule it outside the peak.

Illustration: a bartender places two cocktails on a colleague's tray in a full restaurant while a third staff member walks past with plates

Who stands where during the peak?

An extra pair of hands helps most where the work piles up. So also look at where the extra effort is needed. If guests wait ten minutes for cocktails, an experienced bartender can make more of a difference than an extra person on the floor. And if dishes are ready while the rest of the table is still taking its time, you want to know whether the problem lies with a kitchen station or at the pass. Kitchen lead times are covered in A faster kitchen can mean more revenue.

For each peak hour, write down not only the number of people but also the position: bar, floor, pass, kitchen.

Checklist: test it four Fridays in a row

One evening tells you little. So test a change on four comparable Fridays, in four steps:

  1. Choose one bottleneck: for example, the time between taking the first drinks order and serving it. Write down how you measure it and who does so.

  2. Look at a few comparable Fridays: compare reservations, order times and hours worked side by side. Note any deviations, such as a large group, terrace weather or an event nearby.

  3. Adjust one shift: move a start time, or put someone with the right experience in the busiest position during the peak hour. When drawing up the roster, take contract hours, availability, breaks and agreements on roster changes into account. The rules from the law and the collective labour agreement for this are in Demand-based rostering: plan on expected sales.

  4. Measure the effect: start with the bottleneck from step 1: did the waiting time on those four Fridays get shorter, stay the same or get longer? Then ask the shift leader how the floor, the bar and the kitchen experienced the peak, and whether it still worked at the edges of the evening. Only look at the hours last: were more hours worked than were on the roster?

Common mistakes

  • Only looking at the daily total: wage costs as a share of revenue say something about the whole day, not about the hour in which things went wrong. Which weekly figures to track alongside it is covered in Five figures every hospitality owner must know.

  • Forgetting work without a receipt: if you move the early shift, you may push the preparation into the peak hour.

  • Planning numbers instead of positions: if things jam at pouring and shaking, an extra colleague clearing tables will do little to change that.

  • Judging too early: one quiet or particularly busy Friday says nothing yet about the shift. Wait for the four Fridays to pass before drawing your conclusion.

Do this today

Take your roster for this coming Friday and write down, per hour, how many people are working and in which position: bar, floor, pass and kitchen. Next to it, add what you already know about that evening: the reservations per half hour and, if you have them, last Friday's order times. Then answer two questions. At which hour does your team have the most work? And who is in the position where you need them most at that point? Write down one shift you want to test. Do not change anything in the roster today: this takes at most thirty minutes, and you need no purchase and no Guestit for it.

You will find more articles on staff, kitchen and figures in the Guestit Academy.

Where technology helps

You can collect the three signals with pen and paper. Ask your supplier which of them your system already records. At Guestit, we look at staff planning from the perspective of the whole service. Reservations show when you expect guests: the POS shows today's reservations and the waiting list on one screen. The POS has reports per period, and staff clock in and out on it. The kitchen displays show, among other things, the average ticket time against your target and the slowest station. They do not create a roster. Together, those signals give you a better starting point for your next roster.

How we put this together

This article is a line of reasoning with a worked example, not research. It contains no market figures and no benchmarks for wage costs or revenue per hour: we did not search for benchmarks and therefore cite none. The amount of €4,000 and the Friday roster are made up; the hours and the staffing per hour were checked with a script. The rules on rosters, breaks and on-call work are set out with sources in the article on demand-based rostering and have not been re-checked here.

The starting point is an article by Nory, a software supplier for restaurants, which we retrieved and reviewed on 21 November 2026. The idea of comparing demand per moment with staffing, and of looking back at waiting time and hours worked after a few weeks, is also in their post; we share that idea. Text, worked example and figures have not been copied; their article is based on British rules and wages, and their figures are those of a supplier.

Do you have a correction or a better source? Email hey@guestit.nl. The images for this article are illustrations (AI-generated), not photos of a real venue and not screenshots of Guestit.

Sources

  1. Guestit Academy, Demand-based rostering: plan on expected sales (2026; with the sources for the rules from the law and the collective labour agreement)

  2. Nory, Demand-based scheduling for restaurants: How it saves your profits (12 November 2026; supplier, only the starting point for the topic)

Want to see this work in your business?

Book an introduction and we’ll look together at which of these ideas fit your business. If Guestit isn’t a fit, we’ll say so.

Book an introduction
See the solutions

Shall we give you a call?

Guestit

The operating system for hospitality. Connects every step of the guest journey with the processes behind it.

© 2026 Guestit