Churn Rate vs Retention Rate

Churn rate is the share of customers you lost in a period; retention rate is the share you kept. For the same cohort and period they are exact complements — 6% monthly churn is 94% monthly retention. There is no information in one that is not in the other; the choice is framing, not measurement.

The subtlety is that each comes in a customer and a revenue flavour, and those are not complements of each other. 94% customer retention says how many logos stayed; 90% gross revenue retention says how many dollars stayed. When the two diverge, size is the story — you are keeping the small customers and losing the big ones, or the reverse.

Customer churn rate = customers lost ÷ customers at the start of the period; retention is the remainder. Starting a month with 100 paying customers and losing 6 is 6% churn, 94% retention. Only count customers who could have churned — excluding those added during the period keeps the cohort honest.

Compound them to see what a month really means over a year: 94% monthly retention is 0.94¹² ≈ 48% annual retention. A churn number that looks small monthly can halve the customer base in a year — which is exactly why the retention framing, compounded, is the more honest lens for planning.

Use churn when hunting problems: it is the active number, the one that spikes, the one you set alerts on and assign owners to. A jump from 5% to 8% churn reads as the emergency it is; the same event as a drop from 95% to 92% retention reads like rounding noise.

Use retention when modelling the future: lifespans, LTV and cohort projections all compound the kept fraction, not the lost one. Boards and investors also think in retention — NRR, GRR and cohort curves are all retention-framed. The teams that get this right use both deliberately: churn for operations, retention for economics.