This paper tackles the problem of companies that offer
subscription services, plans or any other recurring method of
payment. In this commercialization model, it is important to keep
the churn rate low b ut to d efine strategies for reduce churn rate it
is needed to identify what are the main reasons for losing
customers . The application of data mining techniques may assist
to find out patterns that can trace the most likely customers to
become churns. Through this research, using the data mining
process, it was possible to identify that, among other factors, the
non utilization of the main system modules and the high default
rates corroborates for customers to become churn.