Interactive computer systems, that is, systems in which users cyclically interact by getting and providing information, have already a widespread and increasing use in all areas of our society. One characteristic of such systems is that the user behavior affects the system behavior and vice-versa. There is strong evidence that much of the user behavior is reactive, that is, the user reacts to the instantaneous conditions at the action time. This paper presents the reactivity concept and describes a framework to model it in interactive systems, in particular Internet-based systems. We analyze an online auction within the framework. Based on eBay data, we identify attributes that affect the winner bidders' behavior, such as the auction time to finish. This paper presents the first findings towards the formal description and understanding of reactivity patterns in an e-commerce application, which will be useful in improving the application and building novel mechanisms.