Predicting upgrade timing for successive product generations: An exponential‐decay proportional hazard model
提出指数衰减比例风险模型预测消费者对连续产品代际的升级时机，发现过往采用和使用模式能预测升级时间，且动态协变量模型优于静态模型。
In the presence of successive product generations, most consumers are repeat buyers who may decide to purchase a future product generation even before its release. Therefore, after a new product generation enters the market, its sales often exhibit a declining pattern, which renders traditional diffusion models unsuitable for characterizing consumers’ decisions on upgrade timing. In this study, we propose an Exponential‐Decay proportional hazard model (Expo‐Decay model) to predict consumers’ time to product upgrade. The Expo‐Decay model is parsimonious, interpretable, and performs better than do existing models. We apply the Expo‐Decay model and three extensions to study consumers’ upgrade behavior for a sports video game series. Empirical results reveal that consumers’ previous adoption and usage patterns can help predict their timing to upgrades. In particular, we find that consumers who have adopted the immediate past generation and those who play games from previous generations more often tend to upgrade earlier, whereas those who specialize in a small subset of game modes tend to upgrade later. Further, we find that complex extensions to the Expo‐Decay model do not lead to better prediction performance than does the baseline Expo‐Decay model, whereas a time‐variant extension that updates the values of covariates over time outperforms the baseline model with static data.