Five Percent Churn a Month Leaves Half the Customers in a Year
Five percent monthly churn sounds close to three percent. Over a year one keeps half the base and the other keeps a third, and the gap widens every month after that.
The Compounding
Churn is the share of customers, or revenue, lost in a period. The trap is treating it as though it adds up.
Five percent monthly churn does not mean sixty percent a year. It means retaining ninety five percent twelve times over, which leaves about fifty four percent of the base. Three percent monthly leaves about sixty nine percent.
Two percentage points of monthly difference becomes fifteen points of annual difference, and the gap keeps widening in year two.
Churn multiplies rather than adding. Any intuition built on adding it up will understate how much damage a seemingly small rate does over a realistic holding period.
Monthly and Annual Are Not Interchangeable
Companies quote whichever framing looks better, and both are technically accurate.
| Monthly churn | Annual retention | Average customer life |
|---|---|---|
| 1% | 89% | ~100 months |
| 3% | 69% | ~33 months |
| 5% | 54% | ~20 months |
| 10% | 28% | ~10 months |
Average customer life is roughly one divided by the churn rate, which is where lifetime value calculations get their time horizon. That relationship is why a small change in churn moves lifetime value so violently: it sits in the denominator.
Customer Churn Versus Revenue Churn
Losing ten percent of customers who each pay very little is not the same as losing ten percent of revenue.
Most businesses churn small customers faster than large ones, so revenue churn is usually lower than customer churn. When revenue churn is the higher of the two, the company is losing its bigger accounts, which is a far more serious signal and one that customer counts alone would hide.
Early Churn and Late Churn
A single average rate blends two distinct failures.
Early churn, in the first weeks or months, generally means the customer was sold something that did not match what they needed, or never got the product working. That is an acquisition and onboarding problem.
Late churn, after a customer has used the product successfully for a year, means the value faded, a competitor arrived, or the customer changed. That is a product and competitive problem.
They have different causes and different fixes, and the blended number points at neither. This is the specific thing cohort tables exist to reveal.
Voluntary and Involuntary
Not all churn is a decision. Involuntary churn happens when a payment fails: an expired card, a declined transaction, an expired billing address.
In consumer subscription businesses this is routinely a large share of total churn, sometimes approaching half. It is also the cheapest to fix, through retry logic, card updater services, and warning customers before expiry.
A company reporting churn without separating the two is reporting a number that mixes a strategy problem with a billing configuration problem.
The Denominator Question
Churn requires a base to divide by, and the base is a choice. Customers at the start of the period is the common convention. Using the average across the period, or the end of period figure, changes the answer.
In a fast growing company the difference is large, because dividing losses by a base swollen with recent arrivals produces a lower rate. A company growing quickly can show falling churn purely from growth, which reverses the moment growth slows and makes the deterioration look sudden when it was arithmetic all along.
The Bottom Line
Churn compounds, so small rate differences produce large divergences over a year and enormous ones over three. Always establish whether a quoted figure is monthly or annual, customer or revenue, and what base it divides by. Split it into early versus late and voluntary versus involuntary, because the blended average describes no actual problem and points at no actual fix.