This user score predicts whether a user will convert for the first time in the next few days.  Commonly used to target users for personalized offers and messages.  Sometimes used to remove ads for likely converters.

This user score represents the statistical probability that the user will churn from the app soon.  Churn can defined as a specific number of non active days, or a specific unsubscribe event.  Commonly used to reduce the amount of free currency given in promotions; and to increase retention among mid-stage users.  Sometimes used to prioritize incoming support tickets.

This score predicts how much a user will spend within a specified number of days.  Can be used to predict virtual and real world currency.  Commonly used to optimize user acquisition efforts by lookalikes.  Also used to determine very early only how AB tests are performing.