Retention, Churn, Reactivation, and Value: Succinct, Technical Definitions

There can be a lot of ambiguity around definitions for user retention and value. Here is the most cogent combination of definitions that we’ve found.

Reactivation, personified [image credit: aircall.io]

Basic Definitions

Boolean Measures

Member Churn: A member has churned if you believe that that member is gone for good. Churn is a boolean value (i.e. yes a customer has churned, or no a customer has not churned) — there is no in between.

Member Retention is the inverse of churn. A member is retained if they have not churned. Retention is also a boolean value — a member is either retained or not retained, with no shades of gray.

Events

Member Reactivation: If a member previously categorized as churned reappears and can no longer be categorized as churned, that member reactivated on the date when they switched from being classified as churned back to being classified as retained.

Numerical Measures

Member Value (sometimes called customer lifetime value) is a measure of how much a member has contributed to a quantifiable objective, whatever you might define that to be. The important contrast here is that two members that are both still retained may have very different values.

Deciding the Threshold for “Churned”

Coming up with a threshold classification for a member becoming classified as churned can be done in a few ways.

A simple way is to look at time intervals between repeated actions taken by a member. Let’s take the example of a video game where we have good data about the time interval between plays. If we know that 95% of intervals are less than 2 weeks, then we can set the threshold that a member who hasn’t played in the last three weeks is deemed churned.

A more robust method for churn thresholding is to benchmark against reactivation. You can say “Let’s look at the members of that cohort who took the important action in time period __x__ relative to cohort start. The proportion of this group previously classified as churned should be less than __y__%.”

A concrete example of the this is “For any cohort, the percentage of cohort members purchasing 12 months into their lifetime that are considered reactivations should be less than 10%.”