Managing trade-offs in call center agent scheduling: methodology and case study

This paper develops a flexible and tractable scheduling methodology that produces near-optimal call center agent schedules while taking into account the costs associated with customer waiting time, customer abandonment, and call center agents. Our methodology combines integer programming (to find a desirable staffing plan for a given total number of agents) and simulation modeling (to evaluate the weekly costs of a given staffing plan). We describe the advantages of this approach over the traditional scheduling method, and test both methods by building schedules based on actual demand and shift data from an actual call center operated by Expedia.com under a variety of cost scenarios. The new scheduling approach not only out-performs the traditional staffing approach in all scenarios examined, it reduces total weekly costs of the call center's existing agent schedule by 8--25%, depending on the scenario.