Using Eye-Tracking Data of Advertisement Viewing Behavior to Predict Customer Churn
Michel Ballings, Dirk Van den Poel · 2013
The purpose of this paper is to assess the feasibility of predicting customer churn using eye-tracking data. The eye movements of 175 respondents were tracked when they were looking at advertisements of three mobile operators. These data are combined with data that indicate whether or not a customer has churned in the one year period following the collection of the eye tracking data. For the analysis we used Random Forest and leave-one-out cross validation. In addition, at each fold we used variable selection for Random Forest. An AUC of 0.598 was obtained. On the eve of the commoditization of eye-tracking hardware this is an especially valuable insight. The findings denote that the upcoming integration of eye-tracking in cell phones can create a viable data source for predictive Customer Relationship Management. The contribution of this paper is that it is the first to use eye-tracking data in a predictive customer intelligence context.