Understanding customer behavior for enhanced marketing strategies in the mobility industry

This work targets different aspects of marketing analytics, exploiting machine learning models applied to behavioral data of customers interactions with the brand. In particular, the focus will be on predictive analytics of purchase propensity for upselling, churn prediction on the customer base considered and computation and prediction of customer lifetime value. These three major topics will be also considered as dimensions to perform a customer segmentation, that in turn, will be completely behavior-based. The consideration of this type of data is related to the implications of the General Data Protection Regulation, but aims at proving insights on how the latter data improves customer understanding, which in turn may help in optimizing business strategies. The business problem at hand, as well as the techniques and the technologies used to extract data are described in the first part. Next, each topic will be treated separately, and will be characterized by the explanation and the comparison of the algorithms subjected to trials, with attention on the compromise between performance and interpretability.