Enter customers at the start of the period and how many left to get churn, retention, average customer lifetime and the twelve-month decay curve.
Churn is the share of customers who leave in a period. Retention is what is left. The interesting number is neither of those – it is average customer lifetime, which is simply one divided by the churn rate. Four percent monthly churn means the average customer stays twenty-five months. Eight percent means twelve and a half. Doubling churn halves the business.
Multiply lifetime by revenue per period and you have lifetime value, which is the ceiling on what you can afford to pay to acquire a customer. Everything in paid acquisition ultimately resolves to this number, which is why a retention problem always shows up first as a media buying problem: the campaigns did not get worse, the customers just stopped being worth as much.
The short version is that retention is one hundred minus churn, and for a single clean period that is exactly right. The longer formula above matters when the period had new signups in it, because counting them as retained turns a leaking business into a growing one on paper.
Start with 1,000 customers, add 180 new ones, end with 1,120. The naive reading is a 112 percent retention rate. The correct reading subtracts the new arrivals first: 1,120 minus 180 is 940, against 1,000 at the start, which is 94 percent retention and 6 percent churn. Sixty customers left and the headcount hid it.
| Monthly churn | Monthly retention | Average lifetime | Retained after 12 months |
|---|---|---|---|
| 1% | 99% | 100 months | 88.6% |
| 2% | 98% | 50 months | 78.5% |
| 3% | 97% | 33 months | 69.4% |
| 5% | 95% | 20 months | 54.0% |
| 7% | 93% | 14 months | 41.9% |
| 10% | 90% | 10 months | 28.2% |
| 15% | 85% | 7 months | 14.2% |
The gap between the second and fourth columns is where most reporting goes wrong. Ninety percent monthly retention sounds close to ninety-five. Over a year the first keeps 28 percent of a cohort and the second keeps 54. The same five points of monthly difference nearly doubles the customer base you still have at Christmas.
Two practical notes. Retention measured on customer counts and retention measured on revenue answer different questions, and a business can lose a quarter of its accounts while growing revenue if the ones that left were small. And retention is only comparable within a cohort: mixing customers who joined in different months into one number smooths away the early drop-off, which is usually the part worth fixing.
| Segment | Typical monthly churn |
|---|---|
| SMB SaaS, self-serve | 3–7% |
| Mid-market SaaS | 1–2% |
| Enterprise SaaS | 0.5–1% |
| Consumer subscription | 5–10% |
| Subscription box, first 90 days | 10–20% |
Cutting churn from 5 percent to 4 percent moves average lifetime from twenty months to twenty-five – a 25 percent rise in lifetime value, achieved without touching a campaign. The same 25 percent gain through media would mean either a quarter more budget at flat efficiency, or a quarter better cost per acquisition, which nobody delivers on request.
This is why retention work outranks bid tuning in most accounts that have both problems. It also changes what you can afford to bid: a higher lifetime value raises the ceiling on acquisition cost, which lets you win auctions your competitors have to walk away from.
Annualising monthly churn by multiplying by twelve.
Annualising monthly churn by multiplying by twelve. Five percent monthly is not sixty percent annual – it is one minus 0.95 to the twelfth power, or 46 percent. The error runs the wrong way, and it runs the wrong way harder as churn rises, which is exactly when people reach for the shortcut.
The second trap is measuring churn on a growing base. Divide leavers by the customers you finished the month with, rather than the ones you started with, and rapid growth will flatter the number indefinitely. Always measure against the opening cohort.