Average customer lifetime is one divided by the churn rate. Five per cent monthly churn gives twenty months; three per cent gives thirty-three; ten per cent gives ten. That reciprocal relationship is why small changes in churn move lifetime value so violently, and why a subscription business with high churn cannot be fixed by acquiring faster.
It is also why the same figure quoted monthly and annually describes very different businesses — 5 per cent a month is about 46 per cent a year, not 60.
Why does a small change matter so much?
Because the reciprocal is steep at low values. Cutting monthly churn from 5 per cent to 4 extends the average lifetime from 20 months to 25 — a 25 per cent increase in lifetime value from a one-point change.
| Monthly churn | Average lifetime | Annual retention |
|---|---|---|
| 2% | 50 months | ~78% |
| 3% | 33 months | ~69% |
| 5% | 20 months | ~54% |
| 8% | 12.5 months | ~37% |
| 10% | 10 months | ~28% |
Going the other way is worse. Moving from 5 to 8 per cent cuts the average lifetime by nearly 40 per cent, and it cuts LTV by the same proportion — which can turn a working LTV to CAC ratio into a loss-making one without any change in acquisition cost.
Where does the plateau come from?
Signups are roughly constant while churn is a percentage, so churn grows with the base until the two balance. A business adding 350 customers a month and losing 2 per cent stops growing at 17,500 — not because acquisition failed, but because the arithmetic converged.
The newsletter growth article describes the same equilibrium for a list, and the mechanism is identical. The ceiling is signups divided by churn rate, and no amount of additional marketing moves it without also moving one of those two figures.
That is the single most useful thing the model tells you: the ceiling is knowable in advance, and it arrives whether or not anyone planned for it.
Is the average lifetime the right figure?
It is a summary of a distribution that is rarely well-behaved. Churn is almost never constant — it is highest in the first weeks and falls sharply for customers who survive them, which means a single rate averages two very different populations.
A cohort curve is the honest view: track each month’s signups separately and plot what fraction remains. The shape usually shows a steep early drop and then a long flat tail, and the flat part is where the business actually lives.
Using one blended churn rate on that shape overstates the loss from long-tenured customers and understates the loss from new ones, which points improvement effort in the wrong direction — at retention campaigns for loyal customers rather than at onboarding.
What is revenue churn versus customer churn?
Two different questions, and they can move in opposite directions. Losing ten small customers and gaining one large one is customer churn up and revenue churn down.
Net revenue retention captures that by including expansion — upgrades and increased usage from existing customers — and it is the figure that can exceed 100 per cent. A business at 110 per cent net revenue retention grows from its existing base alone, which changes the acquisition question entirely.
For a business with no expansion path, customer churn and revenue churn converge and the distinction is academic.
Cancellation timing is worth measuring separately from the rate. Churn concentrated at the first renewal points at pricing or at an onboarding failure; churn spread evenly across tenures points at gradual disengagement, and the two want completely different responses.
What actually reduces it?
Onboarding, mostly, because that is where the steep part of the curve is. A customer who reaches the point of getting value from the product behaves completely differently from one who never did, and the gap between those two populations is most of the early churn.
Involuntary churn is the other addressable share and the most often ignored — failed payments from expired cards. Retry logic and pre-expiry prompts recover a meaningful fraction of it, and it is a systems problem rather than a satisfaction one.
Price and product fit sit underneath both, and neither is fixed by a retention email.
Questions people ask
How do I convert monthly churn to annual? Not by multiplying by twelve. Annual retention is (1 − monthly churn)¹², so 5 per cent monthly leaves about 54 per cent after a year rather than 40.
What is a good churn rate? Entirely dependent on the market and the price point. Consumer subscriptions run far higher than enterprise contracts, and comparing across them is meaningless.
Should cancellations at the end of a term count? Consistently, whichever you choose. Annual contracts churn in a lump at renewal, which makes monthly churn a poor description of them.
Does the ceiling apply to a growing business? It applies whenever signups are flat. Growing signups raise the ceiling; the ratio is what fixes it.
Is a pause option worth offering? Frequently, yes. A pause converts a permanent loss into a temporary one, and the share of paused customers who return is usually far higher than the share of cancelled ones who rejoin.
One reciprocal sets the lifetime, and the lifetime sets everything else. The LTV calculator applies it directly, the CAC calculator gives the other half of the ratio, the newsletter growth calculator shows the same plateau on a list, and the break-even calculator covers the fixed costs the whole thing has to clear.