Initial and repeat purchase logit model for multi-generation technological product markets

Most new product adoption models have focused on single-generation products. Only recently have researchers begun to focus on the importance of analyzing the consumers' purchase demand in multi-generation products. This paper proposes a model that incorporates both initial and repeat purchases, and allows for leap-flogging behavior in multi-generation technological product markets. Whereas most new product adoption models are based on aggregate market sales, the proposed model is estimated and validated on individual purchase data. Within a logistical modeling framework, our model combines purchase incidence (buy/not buy) component and brand (or `generation') choice components for each time period. These model components allow for individual heterogeneity. Purchase probabilities for buyers are captured as a function of purchase history, buyer expectations of future generations, and preferences of the currently available options. The proposed model is quite parsimonious. It requires relatively simple data for estimation. It is empirically tested using individual-level purchase data from an illustrative pilot study in the multi-generation personal computer market. The model fits and explains the individual purchase behavior reasonably well.