The formulas
contribution per order = average order × contribution margin LTV = contribution per order × orders per year × years as a customer ratio = LTV / CAC payback = CAC / (contribution per order × orders per year) × 12 months
If you know annual churn rather than lifetime, the two are reciprocal: 40% churn means the average customer stays 1 / 0.4 = 2.5 years.
Contribution, not revenue
The most common way to get this wrong is to build LTV out of revenue. A customer who spends 80 € an order, three times a year, for two and a half years has a revenue LTV of 600 €. Against a 40 € acquisition cost that is 15:1, and it means nothing.
At a 35% contribution margin the same customer is worth 210 €, and the ratio is 5.25:1 — healthy, but a third of what the revenue figure suggested. At a 15% margin they are worth 90 €, and a 40 € acquisition cost is starting to look expensive rather than cheap.
The margin to use is what is left after goods, marketplace fees, payment processing, fulfilment and returns — everything that varies with the order. Advertising stays out: that is the cost you are measuring against.
The ratio is the wrong headline
3:1 is the number everyone quotes. It is a reasonable floor, but on its own it says nothing about whether you can survive the growth it implies.
Consider two shops with an identical 5:1 ratio. The first recovers its acquisition cost in four months; the second takes twenty-six, because the customer buys once a year. The first can reinvest each customer’s contribution into acquiring the next one and grow from cash flow. The second has to fund two years of acquisition out of capital before the first cohort pays for itself — the same ratio, an entirely different business.
A useful pair of rules: ratio above 3, payback under 12 months. Miss the first and acquisition is not viable. Miss the second and it is viable but you cannot afford to scale it quickly.
Worked example
Average order 80 €, contribution margin 35%, three orders a year, customers stay 2.5 years, 40 € to acquire one:
- contribution per order = 80 × 0.35 = 28 €
- per year = 28 × 3 = 84 €
- LTV = 84 × 2.5 = 210 €
- ratio = 210 / 40 = 5.25 : 1
- payback = 40 / 84 × 12 = 5.7 months
- orders until repaid = 40 / 28 = 1.4
Under a year and a half of orders to break even on acquisition, then two more years of contribution. That is a business that can spend on growth without borrowing.
Where the inputs come from
Average order value — total revenue divided by number of orders, net of VAT. Use a full year if your sales are seasonal.
Orders per year — orders divided by distinct customers over the same period. Most shops discover this is closer to 1.3 than to the 3 they assumed.
Lifetime — the hardest one, and the easiest to inflate. If you have two years of history, do not assume five years of life. Measure what share of last year’s customers bought again this year; one minus that share is your churn.
Acquisition cost — all marketing spend divided by new customers, not all orders. Counting repeat buyers in the denominator is what makes CAC look half its real size.
Frequently asked questions
No. Use net figures throughout. Including VAT inflates both the order value and the LTV while acquisition cost stays the same, which flatters the ratio by roughly the tax rate.
Use one year and treat the result as a floor. A ratio that already works on a single year is a safe conclusion; one that only works over five years is a forecast, not a measurement.
Not necessarily. A ratio of 10:1 usually means you are underspending — there is profitable demand you are not buying. Ratios that high are common just before a competitor notices the same thing.
Partly. On Amazon or Allegro the customer belongs to the platform, repeat purchases are not attributable to you, and the honest lifetime is often one order. That makes the calculation simple and strict: acquisition has to pay back on the first sale.
ROAS judges a campaign on the immediate sale; LTV to CAC judges acquisition over the customer’s life. A campaign below its break-even ROAS can still be right if the customers it brings buy again — but only if you have measured that they do.