Churn rate
Churn rate is the percentage of users who stop using your product during a given period. It is the exact mirror of retention: retention measures who stayed, churn measures who left. Framing the same loss as a positive number — “we lost 8% this month” — often makes the problem feel more urgent and easier to act on than its retention twin.
How it is calculated
For a period, the basic definition is:
churn rate = users lost during the period / users at the start of the period
And the complementary identity every team relies on:
churn rate = 1 − retention rate
What “lost” means depends on your model. For subscriptions, churn is concrete — a cancellation or a failed renewal — so you can track monthly or annual churn precisely. For free products, churn is softer: a user who hasn’t been active for N days is considered churned, which makes the inactivity window a definition you must fix and keep consistent. Revenue churn is a separate, important cut — the share of revenue lost — which can differ sharply from user churn when your high-value users behave differently from the average.
Why it matters
Churn sets the ceiling on growth. If you acquire 10% new users a month but churn 10%, you tread water no matter how much you spend on acquisition. Because the effect compounds, even small reductions in churn dramatically lift lifetime value — a user who stays longer keeps paying longer. In fact, average lifetime is roughly 1 / churn rate, so halving churn doubles expected lifetime. This is why mature teams treat reducing churn as cheaper growth than buying ever more users.
In games and apps
Game churn is brutal early: a large share of players never return after day 1, which is why the retention curve drops steeply at the start. Teams segment churn by cohort, level, channel and platform to find where it spikes — a difficulty wall, a paywall, a buggy build — and watch which behaviors precede it. Subscription apps focus on involuntary churn (failed payments) versus voluntary churn (deliberate cancels), because the fixes are completely different.
In Keentics
Keentics derives churn directly from your raw activity events, so you define the inactivity window and “active” in product terms and get a figure that matches reality, not a sample. You can segment churn by cohort, channel and country, line it up against the retention curve it mirrors, and use user path analysis to see the steps that precede users leaving. Explore cohort analysis for the grouping that makes churn legible.
Related: Active users · ARPPU · ARPU · Attribution