This research, beyond a typology taking into account (1) the abilities or knowledge level of the different users (expert end-users, non-expert end-users, and data processing expert users), which has already been studied abundantly, and (2) the functional dimension (internal staff, retailer, and customer), characterizing the commercial environment in question, attempts to show the interest of a typology based on their behavior. We identify five user types: curious, indecisive, hesitant, resolute, and nosy. To do this, we use an automated assessment method (tracing), integrated into a decision support telematic tool, intended for any and all end-users by means of Minitel (in a marketing information system). Moreover, this method enables us to define a structure ''H'' for developing data communication tools. The main contributions of this research allows us (1) to design and implement a method integrated to a telematic server for observing its users as well as (2) to define a behavioral dimension for end-user analysis. In fact, this constitutes a reference framework which is especially useful to people working with systems assessment and having available such a behavioral typology (obtained by analyzing the automated tracing) enables them to better understand the reactions of the various end-users.
