In marketing, a predictive model is a mathematical tool used to anticipate the future behavior of a customer or prospect. These models are created by analyzing past customer data. For example, a model designed to predict subscription cancellations is based on the characteristics and actions of customers who have already canceled.

Predictive models in marketing often employ sophisticated techniques and leverage environmental data. For instance, SPSS has identified seven types of predictive models: acquisition, cross-selling, upselling, retention, valuation, message personalization, and risk assessment.

A concrete example is the creation of a predictive model using 600 variables to anticipate the sale of a property. The results of this model can then be used to target opportunities for exclusive sales mandates. This approach allows for the optimization of sales efforts and the personalization of interactions with customers based on their specific needs.

In summary, predictive models in marketing are powerful tools that help businesses better understand and anticipate customer behavior, leading to more effective and personalized marketing actions.