The Prediction Research of Cross-purchase Probability Based on Logit Model

摘要

To measure customers' cross-purchase value, first we must make sure whether there will be any cross-buying behavior and the cross-purchase possibility. Drawing on the related research about cross-purchase and customer behavior, the paper explored the factors that affect the customers' cross-purchase probability, and then built the prediction model of cross-purchase probability using binary Logit model. Empirical results showed that the proposed model got better prediction results. The model has certain reference value to improve the measurement of customer cross-purchase value.