Restaurant staffing is matched to demand by forecasting expected volume for each day and daypart, then building the schedule to that forecast instead of to habit 1. Habit-based scheduling, such as automatically putting three cooks on every Tuesday, creates waste or undue stress, while forecast-based scheduling creates control 1.
Inputs to the demand forecast
- Prior year sales for the same week and recent sales trends 1.
- Reservations and large parties 1.
- Holidays and local events such as concerts and sporting events 1.
- Weather patterns, promotions or specials, and seasonal volume shifts 1.
Forecasting demand is difficult in practice because it depends on numerous uncertain factors that are hard or costly to measure, including weather, local events, competitor actions and seasonal patterns 2.
Turning the forecast into staffing
Labor scheduling begins with demand forecasting, and the forecast must then be translated into staffing needs, a stochastic optimization problem with competing objectives and significant uncertainty 2. Many restaurants overspend by staffing the whole shift for the busiest hour, so schedules should be built around peaks 1. Queuing theory takes the same view, encouraging planning around peaks and variability rather than averages 3. In a retail store study, a simulation model generated performance measures such as average wait time, and alternative staffing schedules were then evaluated by altering the number of clerks in each time interval subject to constraints 4.
Checking the result
Labor numbers must always be balanced against service results and guest expectations, and there is no universal ideal labor percentage for restaurants 1. Restaurants that master both forecasting and cost control can improve profitability without sacrificing the guest experience 1.