Unit economics helps a founder ask whether acquiring and serving a customer can sustain the business. Two useful measures are customer lifetime value (LTV) and customer acquisition cost (CAC) payback. Both depend on assumptions; neither is a guarantee of future returns.

Average revenue per account (ARPA) is recurring revenue divided by the relevant number of accounts. Gross margin is the proportion of revenue left after the direct costs of serving customers. Monthly customer churn is the fraction of customers lost in a month. CAC is the sales and marketing cost of acquiring a customer over a defined period.

Use consistent periods and document which costs and which customers you include. Most disagreements about unit economics are not disagreements about the arithmetic; they are disagreements about scope, discovered too late.

The inputs, and the failure each one hides.
CriterionMetricThe numberWhat it hidesRead this first
ARPARecurring revenue ÷ accounts in scope.One very large customer can move it more than a pricing change.
Gross marginRevenue left after the direct cost of serving a customer.How you allocate support, hosting and onboarding decides the number.
Monthly churnThe fraction of customers lost in a month.Averages hide the difference between a cohort that churns and one that does not.
CACSales and marketing cost of acquiring a customer, over a defined period.It can lag the spend that caused it by months, and it can include spend that never acquired anyone.

Suppose a fictional SaaS has monthly ARPA of $100, gross margin of 80%, monthly churn of 5% and CAC of $400. A simple steady-state estimate gives a customer lifetime of 1 ÷ 0.05, or 20 months. LTV is $100 × 0.8 × 20 = $1,600. Monthly gross profit per account is $80, so CAC payback is $400 ÷ $80 = 5 months. The implied LTV:CAC ratio is 4:1.

Founder2 uses deterministic calculations for these measures. Its lifetime calculation is capped at a configured horizon; zero or negative churn uses that horizon. An undefined division, such as payback with zero gross profit, returns no result instead of a meaningful number. That last behaviour is the one worth copying into your own spreadsheet: a tool that returns a number when the denominator is zero is lying on your behalf.

A 4:1 LTV:CAC ratio is not a verdict on the business. It is a screening test — a way of noticing, early, that the money you are spending to win a customer is unlikely to come back, or that it comes back fast enough to be worth accelerating.

Payback is usually the more useful of the two, because it is about time rather than about a modelled future. Five months of payback means the money returns while you still have the founder's attention to spend on it. Twenty-four months of payback means you are funding a pipeline on your own balance sheet, and that is a completely different company.

Averages can conceal differences between customer cohorts. Churn may change over time, customers may expand or contract, acquisition costs may lag, and cash collection does not always match recognised revenue. Small samples make the estimates especially unstable.

Three of these deserve specific mention. Expansion: if a tenth of your customers grow, a flat LTV model will understate the business and a flat payback model will understate how quickly the money returns — and it will do so consistently, which makes the error hard to notice. Lag: if your sales cycle is four months, this month's CAC belongs to revenue you will see next quarter, and comparing it to this month's churn is comparing two different cohorts. And revenue recognition: annual contracts paid upfront look like a terrible month and a spectacular one, which is a cash-flow pattern, not a unit-economics finding.

Compare cohorts and test several scenarios rather than treating one ratio as a verdict. A useful discipline is to recompute the model under three churn assumptions — your best case, your current case and the case where the product loses a third of its customers in a quarter — and to ask which of them you could survive. That last question is the one a ratio cannot answer for you.

The Stripe guide explains how a subscription snapshot can provide useful context while leaving forecasts and assumptions clearly labelled.