On managing operating capacity to reduce congestion in service systems

This paper presents an alternative framework in which service systems are classified according to the ‘visibility level’ of the queue length in the system to the customers. By directly considering the customer's behavior while waiting in the system, we develop two simple models that analyze the benefit of adjusting the ‘operating’ capacity dynamically for two different types of service systems: a system with invisible queue length and a system with visible queue length. The first system is motivated by a customer support service system in which the caller cannot ‘see’ the queue length while the second system is motivated by a check-out service system in which the queue length is highly ‘visible’ to the customers. For each service system, we first specify a policy for adjusting the ‘operating capacity’ dynamically according to the physical environment of the operation. Then we develop a queueing model that analyzes the benefit of managing the capacity dynamically. The analysis provides useful managerial insights for managing capacity in these types of service systems.