Using turndowns to estimate the latent demand in a car rental unconstrained demand forecast

Prior researchers treated turndowns as useless in unconstrained demand forecasts. This study, however, identifies a situation where recorded turndowns successfully predicted the latent demand in a car rental unconstrained demand forecast. It was important to the success of this study to overcome the inherent flaws in the turndown data that inflated latent demand. In this study, a decomposition method that identifies the actual latent demand from the total recorded turndowns is developed. A major US car rental company successfully uses the decomposition method to estimate the latent demand in its yield management system.