What customer lifetime value measures, and what it does not
Customer lifetime value is the total gross margin one customer produces from acquisition until they stop buying. It is the ceiling on what you can rationally spend to acquire that customer, and it is the single number that decides whether a growth plan is a business or a subsidy.
Three qualifiers do most of the work. First, margin, not revenue. Revenue lifetime value flatters every business with a cost of goods; only the gross-margin figure can pay a sales team. Second, gross margin, not net. Overheads that would exist without this customer do not belong in the calculation, but hosting, payment processing, shipping and direct support do. Third, a horizon. The subscription formula quietly assumes an infinite one, which is why the discounted version matters: a dollar of margin arriving in year nine is not worth a dollar today.
Lifetime value is an estimate of a future, so treat it as a range rather than a fact. Its two inputs — revenue per customer and churn — are both measured over a short window and extrapolated a long way. Halving your churn estimate doubles the undiscounted lifetime value, which is why any LTV quoted without the churn figure beside it is unauditable.
Why LTV is a geometric series, and where the division by churn comes from
Start with one cohort of customers and follow their revenue month by month. If a fraction c of revenue disappears each month, then month 0 brings in ARPU, month 1 brings in ARPU × (1 − c), month 2 brings ARPU × (1 − c)2, and so on forever. That is a geometric series with ratio (1 − c), and its sum is ARPU ÷ c. The famous “divide by churn” shortcut is not a rule of thumb; it is the closed form of that sum.
Two consequences follow immediately. The series only converges while c is strictly positive — if expansion revenue matches or exceeds churn, revenue per cohort never declines and no finite lifetime value exists. And the sum is dominated by its tail: at 2% monthly churn, half the total arrives after month 34. Any business that has not existed for 34 months is asserting that tail, not measuring it.
The denominator should be net revenue churn, not logo churn, because it is revenue you are summing. Net revenue churn is gross revenue churn minus expansion from upgrades and seat growth. Logo churn still matters — it sets the average lifespan, which is 1 ÷ logo churn months — but a business whose surviving customers spend more each year has a lower revenue decay rate than its logo churn suggests. Work that figure out with the net revenue retention calculator and your churn rate before you trust an LTV.
Discounting handles the horizon problem. Divide the monthly margin by (d + c) rather than by c alone, where d is your monthly cost of capital, and the far tail shrinks to something you would actually pay for. Gupta and Lehmann showed that this collapses the whole valuation into a simple multiple of current margin — which is why the discounted figure is the one to use in a board pack.
Worked example: $60 ARPU, 3% churn, 0.5% expansion, 75% margin
Take a subscription product billing $60 a month per account, losing 3% of accounts a month, gaining 0.5% of starting MRR back from upgrades, running a 75% gross margin, and discounting at 10% a year.
- Net revenue churn. c = 3.0% − 0.5% = 2.5% = 0.025 per month.
- Monthly gross margin. m = $60 × 0.75 = $45.00.
- Lifetime revenue. $60 ÷ 0.025 = $2,400.
- Gross-margin LTV. $45 ÷ 0.025 = $1,800. Equivalently, 40 months of margin.
- Average lifespan. Using logo churn, 1 ÷ 0.03 = 33.3 months. Note this is shorter than the 40-month revenue multiple, because expansion stretches revenue beyond the average logo life.
- Monthly discount rate. d = 1.101/12 − 1 = 0.0079741.
- Discounted multiple. (1 + 0.0079741) ÷ (0.0079741 + 0.025) = 1.0079741 ÷ 0.0329741 = 30.569.
- Discounted LTV. $45 × 30.569 = $1,375.59.
The discount rate costs you $424 of the $1,800, or 24% of the headline figure. That gap is the price of the assumption that customers keep paying for a decade. Now apply the 3:1 rule: at $1,800 of undiscounted margin you can justify $600 of fully loaded customer acquisition cost; at the discounted $1,375.59 the same rule allows $458.53. Which of those you use is a genuine strategic choice, and you should state which one you used.
How to read the result: ratios, not absolutes
An LTV on its own means nothing. It becomes a decision when you divide it by acquisition cost and when you compare it to the cash you have.
LTV:CAC. The widely quoted benchmark is 3:1 on a gross-margin basis. Below roughly 1:1 you are destroying value with every sale. Between 1:1 and 3:1 the unit economics work but leave little room for overhead or for churn coming in worse than modelled. Far above 3:1 usually means you are underinvesting in growth rather than excelling at it — there is demand you are not buying. Run the number properly in the LTV to CAC ratio calculator.
Payback period. LTV:CAC says nothing about when the cash comes back. A 3:1 ratio with a 30-month payback will bankrupt a company that a 2.5:1 ratio with an eight-month payback would not. The CAC payback period calculator covers that dimension, and for early-stage businesses it is the more urgent of the two.
Consistency of basis. The commonest cause of a wrong ratio is mismatched inputs: a revenue-based LTV divided by a paid-media-only CAC. Both figures must be either gross-margin or revenue, and both must be either blended or paid-only. Mixing them can double the apparent ratio without a single number being individually wrong.
Finally, watch for a rising LTV that comes only from a falling churn estimate. Churn measured over three months on a young cohort is systematically optimistic, because the customers most likely to leave have not had time to leave yet.
LTV as a multiple of monthly gross margin, by net churn
| Net monthly revenue churn | Undiscounted multiple (1 ÷ c) | Discounted at 10%/yr | Discounted at 20%/yr |
|---|---|---|---|
| 1.0% | 100.00 | 56.08 | 40.12 |
| 1.5% | 66.67 | 43.87 | 33.50 |
| 2.0% | 50.00 | 36.03 | 28.75 |
| 2.5% | 40.00 | 30.57 | 25.19 |
| 3.0% | 33.33 | 26.54 | 22.41 |
| 4.0% | 25.00 | 21.01 | 18.36 |
| 5.0% | 20.00 | 17.39 | 15.55 |
Multiples are (1 + d) ÷ (d + c) with d = (1 + annual rate)^(1/12) − 1. The undiscounted column is the same expression at d = 0.
The infinite horizon is doing more work than you think
At 2% net monthly churn the undiscounted formula counts revenue out to month 300 and beyond. Roughly 36% of the total arrives after month 50, which is past the point where most SaaS companies have any cohort data at all. If your company is three years old, a 50-month multiple is a forecast, not a measurement.
Two defences. Use the discounted figure, which prices that tail honestly. Or cap the horizon explicitly: the year-by-year table above shows cumulative margin, so you can quote “three-year LTV” and let the reader see what you excluded. A capped figure that you can defend beats a perpetuity that you cannot.
Mistakes that inflate lifetime value
- Using revenue instead of gross margin. On a 40% margin business this overstates LTV by a factor of 2.5 — and it is the default output of most spreadsheet templates.
- Measuring churn on immature cohorts. Early churn is front-loaded, so a young cohort's monthly rate understates the long-run rate. Use cohorts old enough to have passed their first renewal.
- Mixing annual and monthly rates. Annual churn is not twelve times monthly churn. At 3% monthly, annual survival is 0.9712 = 69.4%, so annual churn is 30.6%, not 36%.
- Netting expansion into churn but not into ARPU. Pick one convention. If you use net revenue churn in the denominator, ARPU must be the current figure, not the initial contract value.
- Averaging across segments that behave differently. Enterprise and self-serve cohorts rarely churn at the same rate. A blended LTV hides the fact that one segment funds the other.
- Charging support and hosting to overhead. Those are cost of revenue for most software businesses, and excluding them is the easiest way to manufacture a 90% gross margin.
- Quoting LTV without the churn assumption. Anyone reading your number needs the denominator to judge it. State the churn rate and the horizon every time.
Choosing between the subscription and retail models
Use the subscription model whenever revenue arrives on a contract that renews until cancelled. Churn is observable, ARPU is a monthly fact, and the perpetuity form is the right shape.
Use the retail model when purchases are episodic: e-commerce, restaurants, professional services on repeat engagements. Here there is no cancellation event to observe, so you substitute two things you can measure — how much a customer spends per order and how often they order — and one thing you must estimate: how long they keep coming back. Get average order value from the average order value calculator and the retention side from the customer retention rate calculator.
The two models are the same equation in different clothes. Retail LTV of AOV × frequency × lifespan × margin is exactly the subscription form with monthly margin replaced by annual margin and 1 ÷ churn replaced by an explicit lifespan. If you know your annual repeat rate, the retail lifespan is 1 ÷ (1 − repeat rate) years, and the two approaches reconcile.
Both are historic models: they extrapolate an average. Probabilistic alternatives such as the Pareto/NBD and BG/NBD families, described in Fader and Hardie's work, estimate each customer's own purchase and dropout rates from their transaction history and produce a distribution rather than a point estimate. Those are the right tools when you have per-customer transaction logs and are deciding who to target. For budgeting, pricing and board reporting, the closed forms on this page are what practitioners actually use — and for revenue per user specifically, the ARPU calculator is the cleaner starting point.
Key terms
- ARPU
- Average revenue per user or account for a period: total recurring revenue divided by the number of accounts. The numerator of the subscription LTV formula.
- Gross churn
- Customers or revenue lost in a period as a share of the starting base, before any offsetting expansion. It sets average lifespan.
- Net revenue churn
- Gross revenue churn minus expansion revenue from existing customers. Negative net revenue churn means the surviving base grows on its own.
- Gross margin
- Revenue minus cost of revenue, as a percent. For software that includes hosting, payment fees and direct customer support.
- Discount rate
- The annual return you require on capital, used to convert future margin into today's money. Venture-stage businesses commonly use a high one.
- LTV:CAC
- Gross-margin lifetime value divided by fully loaded acquisition cost. The conventional target is 3:1; the two figures must share the same basis.