Forecasting Call Arrivals to Call Center of a Telecommunication Call Centre in Sri Lanka

Abstract

Call centers (which is also known as Customer care units) have become one of the increasingly popular topics in today’s society since it helps to develop a good interaction between customers and the company. One of the most important decisions making the task for the call centers is determining the appropriate number of staff officers to meet the demand for ensuring customer satisfaction while minimizing the service cost. For that, it is important to identify the accurate predictions on call arrivals. This study is focused on developing a model for the call arrival data of a major call center of a telecommunication company in Sri Lanka with the purpose of getting precise predictions of hourly call arrival count. The mixed model approach is used in this study for model the call arrival data by incorporating both the fixed and random effects which affect to the call arrival count. For confirming the accuracy of the mixed model, here it included the comparison of the mixed model with the model with only fixed effects. And the mixed model which contains the day of week and hour of the day as fixed effects and daily volume deviation and hourly random deviation as random effects is identified as the precise model for predicting the call arrival count of the call center.