Competing risks for train tickets – An empirical investigation of customer behavior and performance in the railway industry

Based on a comprehensive data set of German railway customers we analyze consumers’ choices and particularly subsequent changes of two-part pricing contracts (loyalty cards). In a competing risks framework, we simultaneously estimate effects on three types of contractual events: cancellations, upgrades, and downgrades. Focusing on customer relationship management (CRM) practices, we find several factors affecting these events, some of which railway companies can influence to their advantage. Intuitively, installing auto-renewal procedures for loyalty cards decreases cancellation hazards. However, automated electronic mailings (e.g., reminders and account statements) and advertising (e.g., ticket offers) can be counterproductive and increase the risk of cancellation.

► We use competing risks models to study consumer choices of 2-part pricing contracts.
► We examine comprehensive travel histories using longitudinal data and 4m transactions.
► We establish what factors influence contract choices, and their effect strength.
► Simultaneously estimated hazards explain contract cancellations, up- and downgrades.
► The model can help to adjust and fine-tune customer loyalty programs.