Calculating LTV when churn is lumpy and contracts renew annually

Key Takeaways

• The standard LTV shortcut, customer lifetime = 1 ÷ churn rate, assumes churn is spread evenly across every period, which understates risk for annual-contract businesses where most churn actually happens in a single lump at each renewal date.
• A business with a mix of gradual and renewal-spike churn should apply a discount adjustment to the simple formula rather than using it unmodified, since blending the two patterns without adjustment distorts the resulting LTV figure.
• Cohort analysis, tracking a group of customers who joined in the same period and watching when they actually stop renewing, is the more reliable method for annual-contract businesses than a churn-rate-derived formula.
• The estimate of customer lifespan should come from observed historical cohort behaviour, when did the majority of past cohorts actually stop renewing, not from a single blended churn-rate assumption applied uniformly.

The formula "lifetime = 1 ÷ churn rate" is clean, and it's also built on an assumption that doesn't hold for a business selling annual contracts: that churn happens continuously, a little bit every month, rather than in a single decision point once a year. A business with 25% annual churn isn't losing a steady 2% of customers every month, it's very likely losing close to 0% for eleven months and then losing a chunk of the book all at once when renewal notices go out. Using the smooth formula on lumpy, renewal-driven churn produces an LTV number that looks precise and is quietly wrong.

Why the standard formula assumes something that isn't true

Customer lifetime, in the basic formula, equals 1 divided by the churn rate, with lifetime and churn measured in matching units (months or years) (Baremetrics, calculating and increasing LTV, retrieved 2026-09-08). A 25% yearly churn rate implies an average customer lifespan of 1 ÷ 0.25 = 4 years. The formula's underlying assumption is that this churn is distributed evenly, in a monthly-billed business, that's often close enough to true; in an annually-billed business, it's usually not, since the actual decision to leave happens once a year, at the renewal point, not gradually across the twelve months in between.

Where the formula breaks for annual-contract businesses

For annual renewals, churn commonly occurs in a mixed pattern: some attrition happens evenly over time (a customer leaving mid-contract, going out of business, changing needs), but a larger spike concentrates at each renewal date specifically (Consero, lifetime value and recurring revenue businesses, retrieved 2026-09-08). If your company's churn contains a mixture of these two patterns, a discount adjustment to the simple LTV formula is needed to account for that variance, rather than applying the unmodified 1 ÷ churn rate calculation (Consero, retrieved 2026-09-08). Practically, this means the naive formula, applied to annual-contract churn data without adjustment, tends to overstate expected lifetime for customers approaching a renewal date, since it implicitly assumes the same steady attrition risk that a monthly-billed business actually experiences, and understates how concentrated the actual risk is at the renewal event itself.

Cohort analysis: the more reliable method for this pattern

Rather than deriving lifespan from a single churn-rate formula, cohort analysis tracks a defined group of customers, everyone who signed in a given period, say, all customers acquired in Q1, and follows what actually happens to that specific group over subsequent renewal cycles (Nudge, cohort analysis for customer lifetime value, retrieved 2026-09-08). The estimate of customer lifespan comes from looking at historical cohorts and observing when the majority actually stop renewing or purchasing, rather than from a formula-derived assumption (Nudge, retrieved 2026-09-08). For an annual-contract business, this means watching what fraction of each cohort survives past renewal 1, renewal 2, renewal 3, and building the LTV estimate from that observed survival curve rather than from a single blended annual churn percentage.

This matters most precisely because the risk is concentrated: a cohort that's 95% intact at month 11 and drops to 78% by month 13 (the renewal window) tells a very different story than a smooth formula assuming even attrition would suggest, and only cohort tracking surfaces that concentration. Run your actual per-cohort renewal data through the CAC/LTV calculator rather than a single blended churn-rate input, since the calculator's output is only as accurate as the churn pattern assumption feeding it.

What this changes about LTV estimates for a services or SaaS business selling annual contracts

For a UAE business selling annual retainers, licences, or subscriptions, this distinction directly affects how aggressively acquisition spend can be justified. A naive LTV estimate that assumes smooth churn will generally overstate expected customer lifetime relative to what a properly discounted or cohort-derived estimate shows, which means a CAC that looked comfortably justified against the naive LTV figure may be less comfortable once the renewal-concentrated reality is modelled in. This is worth revisiting specifically around each major renewal cycle, when a fresh cohort's actual survival rate becomes observable, rather than assuming last year's churn assumption still holds. It connects directly to the broader growth strategy for a services business, where acquisition spend decisions rest on exactly this LTV estimate being realistic.

Frequently asked questions

Can I still use the simple 1 ÷ churn rate formula for an annual-contract business?

You can as a rough starting estimate, but it will tend to overstate expected lifetime because it assumes churn is spread evenly rather than concentrated at renewal. Apply a discount adjustment if you know churn is renewal-spiked, or better, move to cohort-based tracking once you have enough historical renewal data.

How much historical data do I need before cohort analysis becomes reliable?

Ideally at least two to three full renewal cycles per cohort, since a single renewal data point can't distinguish a one-off anomaly from a genuine pattern. A newer business without that history should treat any LTV estimate, cohort-based or formula-based, as provisional and revise it as more renewal cycles complete.

Does this issue apply to monthly-billed businesses too?

Less so. Monthly billing generally does produce more evenly distributed churn, since the "renewal decision" effectively happens every month rather than once a year, which is closer to what the simple formula assumes. The lumpy-churn problem is specifically sharper for annual (or other long-period) contract structures.

The bottom line

The 1 ÷ churn rate formula isn't wrong, it's built on an assumption, smooth attrition, that annual-contract businesses don't actually satisfy. Churn concentrated at renewal needs either a discount adjustment to the simple formula or, more reliably, a cohort-based approach that tracks what actually happens to real customer groups across real renewal cycles, rather than a single blended rate applied uniformly across the year.

Figures and methodology were verified on 8 September 2026 against published customer lifetime value and cohort analysis research. The specific discount adjustment needed depends on your business's actual churn distribution; build LTV estimates from your own historical cohort data rather than a generic formula wherever renewal history is available.