Development of a Wait Time Prediction Model for the Service Industry

Americans do not like to wait in queuing lines, and businesses in the service industry currently do a poor job at forecasting and providing customers with accurate estimated queue times. The service industry needs a better technique to estimate how long customers will be waiting in a queue and a method to easily distribute this information to customers in a timely manner. Current wait time estimating techniques are hardware dependent, and a large-scale implementation involving numerous locations scattered over a wide region may not be feasible for many organizations. Given this, the purpose of this research was to develop a system to accurately estimate customer wait times using predictors that consisted solely of data commonly available on the Internet. Four models were successfully developed that used either regression analysis or an artificial intelligence technique (ANFIS) to achieve this goal. After determining the error in each of these four models, a final model was created that used a combination of the regression analysis and ANFIS to minimize error in the forecasting model. This technology has the potential for widespread adaption in varying types of businesses and organizations throughout the service industry