Unit economics explained: CAC, LTV and payback, with examples

Picture a founder in the middle of a pitch. The investor raises a hand: “What does it cost you to win one customer?” She says €1,500. “And what does that customer leave you with?” She glances at her slides. The slides don’t say.

That second question is the whole subject of unit economics. By the end of this page you’ll know how to work out CAC, LTV, contribution margin, payback and the LTV/CAC ratio. You’ll have followed two invented examples with every step of the arithmetic, and you’ll know where the famous “3 to 1” rule comes from and when to ignore it. New to fundraising? Start from our beginner’s map from zero to a first round.

What is unit economics, and why do investors ask about it?

A unit is whatever you sell one of: a customer, an order, a subscription, a delivery. Unit economics asks whether you make or lose money on that single unit, before rent, before salaries, before everything else. Say a café sells an espresso for €1.20 and spends €0.35 on coffee, water and the cup. It keeps €0.85 a cup, and the rent has to come out of that.

For a startup the question bites harder, because the first sale rarely pays for itself. You spend today to find the customer and collect the margin over months. If the margin never catches up with the spend, revenue can climb every month while the company sinks. CB Insights looked at 431 venture-backed startups that shut down since 2023 and could identify reasons for 385 of them. Running out of capital was named by 70%, which CB Insights calls “almost always the final cause of death, not the root problem”. Unsustainable unit economics was among the reasons for 19%.

Investors ask because the answer tells them whether growth will make your company stronger or weaker. Our guide to what investors look for first in a first-time founder shows where it fits among the other questions.

Unit economics formulas: CAC, LTV, margin and payback

A handful of formulas covers almost everything. They’re in the table, written for a subscription business and counted per month. Two of them deserve a closer look, because that’s where the numbers usually go wrong.

Metric	In plain words	Formula
CAC (customer acquisition cost)	What you spend to win one customer	Sales and marketing cost in a period ÷ new customers won in that period
Contribution margin	What one customer leaves each month after the costs of serving them	Revenue per customer − variable costs per customer
Monthly churn	Share of customers who leave in a month	Customers lost in the month ÷ customers at the start of the month
Customer lifetime	How many months a customer stays, on average	1 ÷ monthly churn
LTV (lifetime value)	What one customer leaves you over their whole lifetime	Contribution margin per month × customer lifetime
LTV/CAC	Euros that come back for each euro spent winning the customer	LTV ÷ CAC
CAC payback	Months until one customer has repaid their own CAC	CAC ÷ contribution margin per month

Start with CAC. a16z says it “should be the full cost of acquiring users”, and it means full. Ad spend counts. So does the pay of the person who runs the demos, the CRM licence, the agency, and the referral fees, credits and discounts that a16z lists among the costs that often go missing.

Then LTV: the margin a customer leaves you across the whole relationship. The right yardstick is contribution margin, meaning the customer’s revenue minus the variable costs that come with them (hosting, payment fees, support, delivery). Not revenue. And not your company’s net margin either, because that has your marketing and salaries inside it, so you’d count acquisition twice. a16z lists “present value of revenue” and even plain gross margin among the common mistakes. The last ingredient is churn, the share of customers who leave each month. Lifetime is 1 divided by churn: at 2% a month, 50 months.

Unit economics example 1: B2B software for clinics

Take a Milan startup that sells booking software to private physiotherapy clinics for €120 a month. It’s invented, like every number in this article. Each clinic costs €24 a month to serve (hosting, payment fees, support), so €96 of the €120 stays. Two clinics in a hundred cancel every month. Last quarter the company spent €60,000 on sales and marketing: a salesperson’s pay, demo software, LinkedIn ads, a stand at a trade fair. It won 40 new clinics.

Step	Calculation	Result
CAC	€60,000 ÷ 40 clinics	€1,500
Contribution margin	€120 − €24	€96 a month
Customer lifetime	1 ÷ 0.02	50 months
LTV	€96 × 50	€4,800
LTV/CAC	€4,800 ÷ €1,500	3.2
CAC payback	€1,500 ÷ €96	15.6 months

Read it the way an investor would. A ratio of 3.2 clears the famous bar. A payback of 15.6 months doesn’t clear the other one. Both facts are true at once.

Now pull one lever. If four clinics in a hundred left each month, lifetime would fall to 25 months, LTV to €2,400 and the ratio to 1.6. Same company, one changed assumption, a different verdict. Skok gives a sense of the scale: at 2% monthly churn, he writes, you lose about 22% of your revenue every year.

And those 50 months are a forecast. If the company started selling a year ago, nobody has watched a clinic stay for four. When data is thin, a16z prefers to measure 12-month and 24-month LTV rather than predict a lifetime. On these numbers the 24-month LTV is about €1,840, or 1.2 times CAC, and more than 60% of the €4,800 sits beyond month 24.

Unit economics example 2: a consumer app

Now a Turin app that sells home cooks a recipe-and-shopping-list subscription (invented too). Subscribers pay €8 a month once IVA is taken out: the price on the screen is higher, but IVA goes to the state, not to you. Payment fees, cloud costs and support eat €2, leaving a margin of €6. Then churn: one subscriber in ten cancels each month, so the average stay is 10 months. €6 times 10 months gives an LTV of €60.

Last month the founders put €5,000 into Instagram ads. 250 paying subscribers came out of it, so paid CAC is €20. Another 250 people subscribed without any ad, through word of mouth and App Store search. So which CAC do you show?

Measure	Paid only (250 subscribers)	Blended (500 subscribers)
CAC	€20	€10
LTV/CAC	3.0	6.0
CAC payback	3.3 months	1.7 months

The blended column looks wonderful. Double the ad budget, though, and nobody promises another 250 free subscribers. a16z writes that investors consider paid CAC more important than blended CAC, and this is why. Show both, side by side, and say which one you’d bet on.

Is a 3:1 LTV/CAC ratio good? Where the rules of thumb come from

The “3 to 1” rule comes out of software-as-a-service. David Skok wrote in 2013 (his page was last modified in June 2026) that the best SaaS businesses have an LTV to CAC ratio “higher than 3, sometimes as high as 7 or 8”, that many of them “recover their CAC in 5-7 months”, and that profitability turns “anemic” once recovery stretches beyond 12 months.

Bessemer Venture Partners repeats the 3x in its 2019 “10 Laws of Cloud” and, in 2021, set payback targets by customer size: under 12 months for small business, under 18 for mid-market, under 24 for enterprise.

Three limits. First, these are benchmarks for subscription software. A marketplace, a hardware company or an online shop has a different shape of margin and repeat purchase, and a rule borrowed from SaaS can mislead. Second, Skok stresses they are “only guidelines” and notes that many healthy SaaS companies miss them in the early days. Third, a ratio can be too good: far above 3, you may be spending too little on growth, and Skok shows how to work backwards from the guidelines to what you can afford to spend.

Payback is a cash question as much as a margin one. A 16-month payback means you finance sixteen months of every new customer before the euro comes back, which is hard to do with six months of runway. Check it against the plan in your startup financial model.

Five mistakes that flatter your unit economics

• Blended CAC. Free sign-ups hide what paid growth really costs. Show paid and blended side by side.
• Leaving costs out of CAC. The founder’s own selling time, the agency, the referral fees, the discounts. If someone is paid to win customers, it goes in.
• An LTV with no churn behind it. Lifetime is 1 divided by churn, so a guess about churn is a guess about everything. What counts as normal depends on what you sell: Lenny Rachitsky and Casey Winters’ 2020 benchmarks put “good” six-month user retention at about 25% for consumer social apps, 40% for consumer subscription software, 60% for small and mid-sized business software and 70% for enterprise software. Your own numbers beat any benchmark, so track them cohort by cohort (a cohort is the group of customers who joined in the same month).
• Revenue where margin belongs. LTV runs on contribution margin. For consumers, strip out IVA first: €1 at the checkout isn’t €1 of revenue.
• One average for everyone. Customers who arrive through a friend’s recommendation aren’t the ones who arrive through ads. Split by channel and by plan before you trust the average.

Your unit economics checklist

1. List every cost of winning customers last quarter, salaries included, and divide it by the new paying customers.
2. Work out contribution margin per customer: revenue (net of IVA) minus the costs that grow with each customer.
3. Monthly churn, measured from real cohorts and never borrowed from a rival’s pitch.
4. Work out LTV, LTV/CAC and payback, then write down which single assumption would break them.
5. Split CAC into paid and blended.
6. Check which of these numbers belong among the metrics investors ask for at seed, and where they sit in the structure of a pitch deck.

Putting a round together in Italy? Our guide to raising capital for a startup in Italy is the hub for this series.