Survival analysis is increasingly used in social science research, and is experiencing a renewal in systems-related research. Here, we present competing risks models, and show how dwell times can be applied to predict users’ online behavior. Unlike Markov chain approaches, survival analysis does not assume independence over time, and it allows for the inclusion of dwell times and covariates that may influence users’ navigation behavior. We model transitions between pages based upon the dwell time of the initial state and then analyze data from a web shop, illustrating how pages that are linked “compete” against each other. The results show significant differences in transition behavior between buyers and non-buyers, which allow organizations to identify potential customers dynamically online. We estimate relative risks for web page transitions based on the dwell time within a clickstream. Our method allows researchers and practitioners to perform clickstream analysis in a detailed manner not previously possible.
|Title of host publication||Oxford Retail Futures Conference|
|Publisher||Oxford Retail Futures Conference|
|Publication status||Published - 2013|
|Event||Oxford Retail Futures Conference 2013 - Oxford, United Kingdom|
Duration: 9 Dec 2013 → 10 Dec 2013
|Conference||Oxford Retail Futures Conference 2013|
|Period||09.12.2013 → 10.12.2013|