Forecasting of Incoming Calls in a Commercial Bank Service Call Center

In this study, we develop forecast models for incoming calls at a call center of a commercial bank in Thailand. We found that incoming calls are non-stationary. Normally, the number of calls is low during holiday and high during the beginning and ending of each month. Various time series models are applied for monthly forecast and an algorithm based on seasonal pattern is proposed for daily forecast. MAPE (Mean Absolute Percentage Error) and RMSE (Root Mean Square Error) are used for comparing the proposed methodology and the current model that the bank uses. The results show that the proposed methodology is better than the current model. MAPE reduces from 9.79% to 8.12% and RMSE reduces from 960.37 to 861.88.