Definition
Lifetime value is the total value a customer generates across their time as a customer. The usual decomposition multiplies average order value by how often the customer buys in a period, by how many periods they stay, and by the gross margin, so that the result is contribution rather than revenue. Each factor is measured on a cohort of customers who started together, because the lifespan of customers acquired last month is unknown and must be inferred from older cohorts.
Two conventions exist. A revenue LTV omits the margin factor and is easier to compute but overstates what the customer is worth; a margin (or contribution) LTV is what can be compared with acquisition cost. Some teams also discount future value to the present, since margin next year is worth less than margin now. State which convention a figure uses, because a "lifetime value" without that label is not comparable with anything.
Avg order value × purchase frequency × lifespan × margin(currency)
How to interpret it
Direction is "higher is better", with the reminder that LTV is an estimate about the future, not a measurement of the past. A historical LTV counts what older cohorts actually produced; a predictive LTV models what newer cohorts are expected to produce. The historical figure is accurate and stale; the predictive one is current and uncertain. Both have their place, and the gap between them is itself a signal about whether retention is changing.
The number is read against what it cost to acquire the customer. If margin LTV is well above CACCost to acquire one new customer., the business can afford to spend more on acquisition; if it is close to or below it, growth is destroying value. The payback period, the number of months of margin it takes to recover acquisition cost, says whether the business can afford the wait even when the lifetime ratio looks healthy. The LTV and CAC economics guide sets out that comparison in full, and the customer analytics route covers how to build cohorts in the analytics stack.
What moves it
- Retention. Lifespan is the factor most sensitive to small changes and the one marketing influences least directly; the retention section covers the levers.
- Purchase frequency. Replenishment products and subscriptions raise it; one-off purchases cap it.
- Order value, through range, pricing and bundling.
- Margin. Discount-led acquisition can raise revenue LTV while lowering margin LTV.
- Acquisition channel. Customers from different sources have different lifetimes, so LTV should be computed per channel where cohorts allow.
- Cohort age. Early figures understate lifespan because the cohort has not had time to churn or to repeat.
- Returns and service cost, which some models subtract and others ignore.
Retention and frequency compound, so a small lift in either moves LTV more than an equivalent lift in order value; margin, by contrast, scales the whole figure at once.
Reading it in context
| Comparison | What it tells you | Watch for |
|---|---|---|
| Margin LTV against CAC | Whether acquisition creates value | Revenue LTV used by mistake |
| LTV by acquisition channel | Which channels bring customers who stay | Small cohorts, noisy estimates |
| LTV by first product | Which entry products lead to loyal customers | Confounding with channel |
| Historical versus predictive LTV | Whether retention is improving or decaying | Model assumptions |
| Payback months beside LTV | Whether growth is affordable in cash terms | A high ratio with a long payback |
Building a margin LTV from a cohort
A coffee subscription business takes the 2,000 customers acquired in one quarter. Their average order is £28, they order 1.6 times a month, the cohort's observed average lifespan projects to 14 months, and gross margin after shipping is 45%. Revenue LTV is 28 × 1.6 × 14 = £627; margin LTV is £627 × 0.45 = £282. Blended CAC for the quarter was £95, a margin ratio of about 3.0 and a payback of roughly 4.7 months. A paid-social cohort from the same quarter shows a 9-month lifespan and a CAC of £120, a ratio of 1.5, which changes where the next quarter's budget goes.
Related metrics
CACCost to acquire one new customer. is the number LTV exists to be compared with; neither means much alone, and the pair should be computed on the same cohort and the same margin basis. AOVAverage revenue per order. is the first factor in the LTV formula and the one most quickly changed by merchandising, which makes it the fastest route to a higher lifetime value when retention is slow to move. ROIProfit relative to cost. is the campaign-level version of the same economics: a lifetime-window ROI uses LTV as its gain, and a first-order ROI ignores most of it. Purchase frequency and retention rate are the other two inputs, usually tracked as cohort curves rather than single figures, because a single average hides the point at which most customers lapse. The inputs and CAC belong on one page of any report.
For a bounded version of lifetime value in practice, the break-even calculator guide for subscription businesses uses revenue within a payback window to set the maximum acquisition cost.
Common mistakes
- Comparing revenue LTV with CAC. Only margin can repay acquisition spend, and the revenue figure makes unprofitable growth look healthy.
- Using the average customer when the distribution is skewed by a few heavy buyers who carry the mean.
- Projecting lifespan from a young cohort, which has not yet had time to churn and so always looks loyal.
- Treating LTV as a fixed constant in budget models while retention drifts underneath it.
- Ignoring channel and product differences that make a blended LTV misleading for allocation decisions.
- Discounting nothing, so that margin expected years from now is counted as though it were already in the bank.
- Reporting LTV without the cohort and window it was measured on, which makes the figure impossible to reproduce or compare with last year's.
