Customer retention measures how many of your existing customers keep buying from you over a defined period. You express it as a percentage: customers still active at the end of the period, divided by customers you had at the start, excluding new ones. High retention means less pressure to win replacements.
You calculate retention over a fixed window, usually a month or a year. Take the customers you had on day one, then count how many still pay on the last day. New customers won during the period stay out of the calculation, because mixing them in hides the losses.
An example. You start January with 400 subscribers. During the month 40 cancel and 60 new ones sign up. Your retention rate is 360 divided by 400, which is 90 percent. Your churn rate is the other 10 percent. You end the month with 420 customers, but that growth number says nothing about whether the product holds people.
Two mistakes come up often. The first is counting the 60 new subscribers in the numerator, which produces 420 divided by 400 and a nonsense rate above 100 percent. Revenue retention can pass 100 percent when existing customers upgrade, but customer retention cannot. The second mistake is quoting a retention rate without a time window. Ninety percent per month and ninety percent per year are very different: at 90 percent monthly, you keep about 28 percent of a group after twelve months, because 0.9 to the power of twelve is roughly 0.28.
Retention drives your other numbers. If customers stay 10 months on average instead of 5, each customer is worth twice as much, and you can afford to pay twice as much to win one.
