Customer Lifetime Value Calculator

Written out

Lifetime value is contribution per order × purchases per year × years retained. The contribution version, not revenue; is the only one that can be compared against acquisition cost.

LTV = average order value × contribution margin × purchase frequency × customer lifespan.

For a subscription business the equivalent form is monthly contribution ÷ monthly churn rate, which is the same calculation expressed through churn rather than through an observed lifespan.

Running the numbers

$58 average order at a 55% margin is $31.90 of contribution. At 2.4 purchases a year over 2.2 years, that is 5.28 orders and $168.43 of lifetime value.

Against a $27 CAC that is a ratio of 6.2 to 1, which is comfortable.

The revenue version of the same calculation, $58 × 2.4 × 2.2, gives $306.24, which is not a lifetime value in any useful sense because most of it is cost of goods.

Businesses comparing a revenue LTV against a CAC are comparing a gross figure against a net one and will conclude they can afford roughly twice what they actually can.

What gets missed

Lifespan is the least reliable input and the one the answer is most sensitive to. A business three years old cannot observe a five-year lifespan and any figure beyond its own history is an extrapolation.

Averages also conceal a steep distribution: a small share of customers generate most of the value, and an average LTV describes neither them nor the majority who buy once and leave.

What to do next

Use contribution rather than revenue, and use a horizon you can observe, twelve or twenty-four months rather than a theoretical lifetime. A twelve-month LTV is verifiable and a five-year one is a forecast.

Then calculate it by cohort and by acquisition channel. Customers acquired through referral and through discount campaigns have very different lifetime values, and the blended figure funds acquisition of the wrong kind.

Why a shorter horizon is usually the better number

A long-horizon LTV justifies a high acquisition cost and pushes the cash return years out, which a self-funded business cannot survive however correct the arithmetic proves to be.

A twelve-month LTV is conservative, checkable against actual cohort data, and produces an acquisition budget the business can fund from its own cash flow.

The businesses that get into trouble with LTV are generally those that used a five-year figure to justify spending today. Being right about the eventual value is no defence if the money runs out in month eight, which is why payback period usually matters more than lifetime value for anyone without external funding.

Discounting matters once the horizon extends beyond a year or two, since contribution arriving in year three is worth less than the same amount today. Applying even a modest rate makes long-horizon figures considerably more defensible.

Cohort-based calculation also reveals whether lifetime value is improving, which an aggregate figure cannot show, a business whose newer cohorts are worth less than its older ones has a problem that the blended number will hide for years.

Predictive models exist that estimate value from early behaviour rather than waiting for the relationship to end, and even a crude version, value at ninety days as a predictor of value at two years, beats an average applied to everyone.