Churn prediction is forecasting which customers are likely to stop buying, stop engaging, or cancel.
Category
AI
Also known as
churn model
It is a practical, high-value CDP and AI use case: catch at-risk customers early enough to act. The usual inputs are recent purchase frequency, declining engagement, support issues, subscription status, and time since last activity.
In practice
A churn model is only worth building if a retention action follows it. Wire the score to a journey, or it predicts loss without preventing it.
Something out of date?