The Effect of Cross-Training and Scheduling in an Inbound Call Center Using Simulation

This thesis presents an analysis of the benefits of cross-training between the claims division and calls division in a large health insurance call center by building a discrete-event simulation model of the call center. The simulation model was built in Rockwell Software's Arena v. 7.0 using a modified trace-style simulation. The simulation is directly driven by eleven months of data in half-hour interval summaries; the randomness of the events that occur within these half-hour intervals has been interpreted so as to fit the average speed of answer to that given in the data.The model of the actual system is then modified to incorporate cross-training. The effect on average speed of answer and claims output of cross-training claims agents that can answer phones when needed is analyzed as well as the effect of cross-training calls agents to process claims when call volumes are low. In addition, the total number of call center agents needed when cross-training is introduced is also examined. The schedules of the call center agents are then manipulated in an attempt to lower the average speed of answer during the busiest periods of the day.The analysis shows that cross-training holds the potential to not only reduce the average speed of answer and increase the rate at which claims are processed, but can also significantly reduce the wage-related costs. Also, simply changing the schedules of the call center agents without adding resource capacity had a great positive impact on the average speed of answer, lowering it immensely during the periods which formerly had the highest average speed of answer. This result is an indication that carefully planned scheduling, along with strong schedule adherence, can have a significant positive effect on call center performance.