Supporting customer retention through real-time monitoring of individual web usage

Abstract

Customer retention is crucial for any company relying on a regular client base. One way to approach this problem is to analyse actual user behaviour and take proper actions based on the outcome. Identifying increased or decreased customer activity on time may help on keeping customers active or on retaining defecting customers. Activity statistics can also be used to target and activate passive customers. Web servers of online services track user interaction seamlessly. We use this data, and provide methods, to detect changes real-time in online individual activity and to give measures of conformity of current, changed activities to past behaviour. We confirm our approach by an extensive evaluation based both on synthetic and real-world activity data. Our real-world dataset includes 5,000 customers of an online investment bank collected over 3 years. Our methods can be used, but are not limited to, trigger actions for customer retention on any web usage data with sound user identification. © 2008 Springer-Verlag Berlin Heidelberg.