Creating a workforce schedule that ensures appropriate service levels is a key management function. The many complexities of scheduling can be captured through a process that comprises the following four major steps: (1) forecasting demand, (2) translating the demand forecast into employee requirements, (3) scheduling the employees, and (4) controlling the schedule as the day unfolds. Each of those steps involves its own set of tasks. To create a forecast, a manager must determine what needs to be done to meet the expected demand for a given planning period. While a planning period may be of any duration, a 15-minute period is an effective one to use. In particular, the manager must identify the demand drivers and assess whether they are time variant (that is, variable over short periods) or time invariant (relatively stable over short periods). Another part of the forecasting step is determining the tasks to be done in a given period. Some of the tasks (notably, those involving direct customer service) are uncontrollable, because they must be done on the spot. Other tasks, though, such as side work, are controllable because they can be performed at any time (within reason). Having created a fairly reliable estimate of demand, the manager must next translate that demand into the number of workers needed, using an economics-based labor standard. At this point, the manager is ready to construct a schedule that will do the best job of deploying the staff to achieve the desired economic standards without overstaffing and inflating costs. Scheduling is subject to hard constraints, or factors that must be addressed (such as the number of hours an employee can work in a day), and soft constraints, or factors that are desirable in a schedule but not essential (such as employees’ desires for when they work and what tasks they perform). Having created a schedule that will meet the economic standards within the constraints, a manager must finally monitor and fine tune the schedule as the day goes on. Most critically, the manager must decide early on whether the demand estimate for the day is correct—meaning the staffing levels will be sufficient—or whether the actual demand is different from the estimate. If the demand estimate proves incorrect, the manager must further decide whether to take such long-lived actions as calling in workers to take care of a big day (or send them home if business has died off ) or merely take a short-lived action (such as sending employees on break) to account for momentary fluctuations in actual demand. Computer applications can assist managers in most of the workforce-scheduling tasks, but a manager needs to understand the process if only to judge whether the application in question is providing solutions that are reasonably close to the optimal schedule.
Thompson, G. M. (2004). Workforce scheduling: A guide for the hospitality industry [Electronic article]. Cornell Hospitality Reports, 4(6), 7-55.