Marketing monitoring systems
Client marketing analysis, in terms of propensity and segmentation models, enrich the client’s information wealth and assist the Marketing and Customer Base divisions in creating front-end activities. This additional information is a key element during contact with clients, both outbound and inbound, and in particular it guides the preparation of specific campaign lists, be they either churn prevention or the supply of various services/products.
The impact that these activities have over time in terms of the reduction of the abandonment phenomenon and earnings from various promotions are of fundamental importance. The proper functioning of the various predictive models must also be checked at the same time, in particular by observing the Customer Base dynamics compared to the phenomena correlated to the same models.
The objectives of designing and creating a monitoring system in a context like this are to gather data and inform in a concise manner on everything that relates to the modelling which is developed and the activities connected to it. Our role, then, is to prepare an automised Reporting system that considers the various aspects to monitor, like, for example:
- transition matrices for studying migrations over time amongst the various behavioural segments in which the customers are subdivided
- campaign redemption analysis compared with test and control samples
- investigated phenomena analysis (churn, propensity, etc.) via a comparison of observed data and predictions supplied by statistical models
- investigated phenomena trend analysis
- investigated phenomena analysis compared to customer characteristics
- statistical model performance monitoring in terms of maintaining the discriminant capacity and customer profiles in relation to the various phenomena propensity.