Rule-based and data-driven action models are often differentiated in online business. In general, rule-based here means static, because a fixed model and rules that can be defined by the business user or based on predefined user-persona are used. Accordingly, each digital touchpoint of a user profile is assigned a specific follow-up action at certain points along the customer journey. Thus, there is no flexible process. A system is rule-based when data is organized exclusively via hypotheses in the form of predetermined rules. As an example, the following can be given: “The third call of an advertisement shows clear buying interest”.
Data-driven models, on the other hand, are dynamic and recalculate the follow-up actions of each click along the user’s customer journey in a completely individual way. Evaluation and implementation take place permanently and in real time. In background processes, several click paths are compared, whereby both conversion-strong and aborted paths are taken into account. Data-driven is defined here, therefore, by the use of underlying data in order to recognize certain patterns of causality within seconds and to spontaneously transfer them into a dynamic model. In an online shop, for example, it is possible to show previously completely unknown users without any visitor history situational product suggestions based on the access device or even the current weather.