Subscription Retention Calculator

How the number is derived

Retention rate is the complement of churn: subscribers remaining at the end of a period ÷ subscribers at the start × 100, excluding new additions.

Cohort retention tracks a single group over time and produces a curve rather than a rate, which is far more informative.

Logo retention counts subscribers; revenue retention counts money, and the two diverge whenever departures are not average-sized.

An example

96% monthly retention compounds to 61% at twelve months: 2,000 subscribers becoming 1,220 if none are replaced.

The cohort curve is usually steeper early: perhaps 88% at month one, 94% each month after. That gives 88% × 0.94^11 = 44% at twelve months, well below the flat-rate projection.

Those two numbers, 61% and 44%, produce lifetime values differing by roughly a third, and the second is the honest one for a business with front-loaded churn.

A cohort curve that flattens at 40% means four in ten subscribers are effectively permanent, which is a different and much better business than one where the decline continues.

The catch

A single retention rate averaged across cohorts of different ages mixes populations behaving very differently, and it moves when the growth rate changes rather than when retention does.

Fast-growing businesses look worse on aggregate retention purely because a larger share of their base is new and new subscribers churn hardest.

What this changes

Always report by cohort at a fixed age: month 1, month 3, month 12, so cohorts are compared at the same point in their life.

Then watch whether the curve flattens and where. The flattening point identifies the durable core, and the size of that core is what determines the business's long-term value.

Why month one dominates everything downstream

The largest single drop in almost every subscription retention curve is between signup and the first renewal. Subscribers who survive it behave completely differently from those who do not.

That makes onboarding the highest-leverage work available, getting the subscriber to the point where the product has delivered something before the first charge arrives.

It also means a business measuring only steady-state churn is measuring the survivors and missing where most of the loss occurs. Splitting the figure into first-period churn and mature-base churn usually reveals a healthier core and a much more specific problem than the blended rate suggests.

Reactivation of lapsed subscribers is worth tracking separately, since returning subscribers behave differently from new ones and typically retain better than first-time signups.

Counting them as new acquisitions overstates acquisition performance and understates how much value the existing relationship still holds.

Comparing retention curves between acquisition channels frequently shows a wider spread than any product change produces, which makes channel mix a retention lever as well as an acquisition one.

Publishing the cohort curve internally, updated monthly, keeps retention visible to people whose decisions affect it and who would otherwise never see the number.

Comparing the curve against the previous year's cohorts at the same age shows whether product and onboarding changes have actually improved retention or merely moved the aggregate.