Modeling Customer Churn with Logistic Regression
Customer Churn is a measure of the percentage of a company's customer base that ceases to be an active customer in a given time period.Probability of Exit
A common method of forecasting customer churn is to estimate the probability that a given customer will exit the relationship in a given time period. A naive measure would be to take the historical churn as the probably of exit:
{% Prob = \frac{\# \; Lost \; Customers}{\# \; Customers} %}
However, it is often desirable to try to understand the factors that influence churn.
Logistic regression
is a common way to estimate these relationships.