Live October 1st. Join waitlist, get early-bird rate for 6 months.Claim yours today →
LTV from your actual retention curve, rather than from a churn rate that pretends to be constant.
The standard LTV formula divides revenue by churn, which assumes churn is constant. It is not — cohorts churn hardest in the first months and then flatten. This uses your real curve instead.
Take these from a real cohort that has aged at least twelve months. Beyond month 12 the curve is extrapolated at the observed month 6–12 decay rate, and capped at five years.
The constant-churn formula understates you here, because your early retention is unusually strong relative to the long tail.
Modelled retention curve
Customers who leave in month one were never a good fit. Customers still there at month twelve have integrated the product into how they work. Those two groups do not churn at the same rate, and averaging them into a single figure loses the distinction entirely.
| Month | Still active | Implied monthly churn |
|---|---|---|
| 1 | 85% | 15.0% |
| 3 | 72% | 6.3% |
| 6 | 64% | 3.2% |
| 12 | 58% | 1.6% |
| 24 | 50% | 1.2% |
Implied churn falls from 15% to under 2%. Applying the month-one rate to the whole lifetime would value this cohort at a fraction of its real worth — and applying the month-24 rate would wildly overvalue it.
Most SaaS cohorts stabilise somewhere between months 6 and 12. That flattening point is the most useful thing on the chart: it tells you when a customer has genuinely adopted the product, which is the moment onboarding should be aiming at.
Compare the result here against the LTV calculator. If the cohort figure is much higher, your early churn is masking genuinely durable long-term customers. If it is much lower, the flat tail you assumed does not exist.
GOOD QUESTIONS
From a cohort that signed up at least twelve months ago. Take everyone who joined in a single month and count how many were still paying at months 1, 3, 6 and 12. Most billing platforms and analytics tools produce this directly.
Because 1/churn assumes a constant rate. Real cohorts churn hardest early and then flatten, so a single early rate applied forever understates the customers who survive — often substantially.
That is a meaningful signal. A curve declining steadily with no plateau suggests the product does not become more valuable with use, so customers eventually leave regardless of tenure. It is a product problem, not an onboarding one.
Five years is a reasonable ceiling and is what this uses. Beyond that the tail contributes little in discounted terms and the uncertainty is large. For planning cash, the first-year figure is far more useful than the lifetime total.