Forecasting time-varying arrivals: Impact of direct response advertising on call center performance

•We analyze high-frequency advertising and call data by specifically addressing subtle seasonal cycles.•We assess the impact of advertising on call center performance.•We develop a forecasting model for the incoming calls.•We create a discrete event simulation to simulate the manpower planning system.•We demonstrate the importance of the link between marketing activities and operational management. This study investigates manpower planning and the performance of a national call center for scheduling car repairs and responding to road interventions. We model the impact of advertising on the required capacity and develop a forecasting model for incoming calls, where the impact of direct-response advertising is considered. With the estimation results, we forecast the number of incoming calls to the call center. Next, the forecasts are input into the capacity planning simulation module to directly simulate a service process at the highly disaggregated level. This simulation mimics the service level requirements and queue behavior and shows that the call center is operating at a high level of efficiency and performance. We illustrate that advertising may cause a temporary overload of the system and increase the number of abandoned calls, which is suboptimal for call center performance.