Skip to content
Log in
← Statistics glossary

retention

Also called: retention rate, customer retention

The share of customers (or revenue) that stays over a period; for one month it equals 1 − churn. High retention is what makes a subscription business grow instead of refilling a leaking bucket.

Customer retention is the share of customers who are still paying at the end of a period out of those who were paying at its start. For one month it is simply 1 − churn. The usual formula: retention rate = (customers at the end − new customers acquired during the period) ÷ customers at the start.

For a subscription business, retention decides whether growth is possible at all. Every new customer is poured into a bucket; churn is the hole in it. With high retention, each month's sales add to the base; with low retention, sales mostly replace customers who just left. Revenue retention (how much of last period's revenue the same customers still bring) is the money version of the same idea and is covered by net revenue retention.

Retention is earned in the first weeks. Customers who get the product set up and see a result early (in a reminder app, the first customer who comes back after a reminder) stay far more often than those who sign up and drift. So the main levers are onboarding, reaching that first result fast, and removing the reasons people leave that you can fix, such as failed payments, rather than discounts at the moment of cancellation.

Example

GarageDesk starts October with 58 paying shops, wins 9 new ones and ends the month with 64.

Retention = (64 − 9) / 58 = 55 / 58 ≈ 94.8%, which matches churn of 3 / 58 ≈ 5.2%. Retention is decided early: a shop that never imports its client list never sees a car come back after a reminder, and those are the shops most likely to leave. So the founders treat that first returning car as the moment a shop is activated and follow up with every new shop that hasn't reached it yet.

Common mistakes

  • Counting new customers as retained. Subtract customers won during the period before dividing.
  • Reading retention without the reasons. Talk to (or at least list) every customer who leaves; the pattern matters more than the percentage.
  • Fighting churn only at the cancel button. Discounts at exit keep a few customers; faster onboarding keeps many.
  • Looking only at logos. Keep an eye on revenue retention too, since large customers weigh more.