An Empirical Comparison of Weibull and Its Extended Models for New Product Trial Calibration and Forecasting

Weibull distribution is a convenient and flexible model for new product trial calibration and forecasting, however, the impacts of ´never triers´ and heterogeneity in consumers are largely neglected in previous studies. In this study, we extend Weibull model by incorporating these two impacts, and make an empirical comparison across Weibull and its extended models. The results reveal that calibration and forecasting accuracy of the extended models are significant improved.