Unit economics measures the profit on one unit of your business — a single customer, order, or subscription — and the target most software operators aim for is an LTV:CAC ratio of 3:1. Calculate it in three moves: find the contribution margin (price minus variable cost per unit), calculate customer acquisition cost, then estimate lifetime value. If LTV comfortably clears CAC and you earn CAC back within about 12 months, the unit is healthy. Below a 1:1 ratio, every sale loses money.
What Is Unit Economics, Exactly?
Unit economics is the revenue and cost tied to one single unit of your business. A "unit" is whatever you sell repeatedly — one customer, one subscription, one order, or one delivery.
The core number is contribution margin: unit price minus the variable cost of delivering that unit. Variable costs scale with each sale (payment fees, hosting per user, support time). Fixed costs — rent, salaries, your own time — sit outside the unit calculation.
Positive unit economics means each sale contributes cash toward fixed costs and profit. Negative unit economics means you lose money on every unit and scale makes the hole deeper. This is why unit economics matters before you scale: growth only helps a business whose units already work, and any owner spending their own money wants proof that one unit makes money on its own.
How Do You Calculate Unit Economics Step by Step?
Calculate unit economics in four steps: define the unit, find its contribution margin, calculate CAC, then estimate LTV. Contribution margin is price minus variable cost per unit; every other number builds on it.
You might also like
- Define the unit. For SaaS, use one customer or one account. For e-commerce, one order.
- Contribution margin. Subtract variable cost from price. Example: a $50/month plan costs $8/month to serve (hosting, Stripe fees, support), so the margin is $42.
- Customer acquisition cost (CAC). Divide total sales and marketing spend by new customers won. Spend $1,800 to win 10 customers, and CAC is $180.
- Lifetime value (LTV). Multiply average monthly revenue by gross margin, then divide by monthly churn. At 4% churn, the average customer stays 25 months, so LTV is $50 × 0.84 × 25 = $1,050.
In this example the LTV:CAC ratio is $1,050 ÷ $180 = 5.8:1, and CAC payback is $180 ÷ $42 = 4.3 months. Both clear the healthy thresholds.
What Are CAC and LTV, and How Do They Fit Together?
CAC is what you pay to acquire one customer; LTV is the gross profit that customer generates before they leave. Together they tell you whether acquisition is an investment or a leak.
Andreessen Horowitz's breakdown of 16 startup metrics defines the pair as the backbone of any subscription model. If a customer costs $180 to acquire and returns $1,050 in gross profit, acquisition is an investment. If they cost $1,050 and return $180, it is a slow leak dressed up as growth.
| Metric | Formula | Healthy benchmark |
|---|---|---|
| Contribution margin | Price − variable cost/unit | Positive, ideally >50% of price |
| CAC | Sales & marketing spend ÷ new customers | Recovered within 12 months |
| LTV | ARPA × gross margin ÷ churn rate | ≥ 3× CAC |
| LTV:CAC ratio | LTV ÷ CAC | ~3:1 |
| CAC payback | CAC ÷ (monthly ARPA × gross margin) | < 12 months |
The 3:1 ratio and the sub-12-month payback are benchmarks popularized in David Skok's SaaS metrics benchmarks at For Entrepreneurs. A ratio far above 3:1 often means you are underspending on growth, not winning.
How We Track Unit Economics Shipping Software Every Day
We build and ship production software with AI daily, and here is what actually moves our unit economics: variable cost per user is no longer just hosting — it is LLM API calls. When we shipped an AI feature that called a large model on every request, our per-user variable cost tripled overnight, and a plan that looked 84% gross margin quietly dropped toward 60%.
We fixed it the way any owner should: measured the real cost per active user, cached responses, and routed cheap requests to a smaller model. Contribution margin recovered without raising the price. We now check cost-per-active-user in the same weekly review as revenue — one spreadsheet, updated by hand, because the number is too important to outsource.
The lesson we relearn constantly: own the number, and update it whenever the stack changes. Renting a dashboard that reports blended CAC hides the one input — model cost per user — that decides whether the feature is profitable. Bessemer Venture Partners' Cloud economics atlas makes the same case: efficiency, not raw growth, is what compounds.
Which Numbers Actually Matter (and Which Lie)?
The numbers that matter are paid CAC, gross-profit LTV, and cohort churn; the ones that lie are blended CAC and revenue-based LTV. Blended CAC divides all spend by all customers, hiding how expensive paid acquisition really is.
Two traps to watch:
- Revenue LTV instead of gross-profit LTV. Using top-line revenue ignores the cost to serve and inflates LTV by whatever your gross margin is missing.
- Averaged churn. New cohorts churn faster than old ones. Blend them and you overstate customer lifetime, which overstates LTV.
Payback period matters as much as the ratio. A 4:1 LTV:CAC with a 30-month payback still starves a bootstrapped business of cash long before that value arrives. Always separate organic from paid, because a low blended CAC can mask paid channels that lose money on every click.
Common Mistakes That Break Your Unit Economics
The most common mistake is leaving costs out of the unit — support, payment fees, and per-user infrastructure all belong in variable cost. Skip them and every downstream number is wrong.
- Forgetting variable costs like Stripe fees, LLM tokens, and support labor.
- Counting revenue as LTV instead of gross profit.
- Using lifetime = 1 ÷ churn without a gross margin haircut.
- Ignoring CAC payback and running out of cash while "profitable" on paper.
- Optimizing LTV:CAC past 5:1 by starving growth.
Fix the inputs first. Unit economics is only as honest as the costs you are willing to count.
0 Comments
Log in to comment
Not a member yet? Join the community
Pick a meme
KlipyHave a great take?
Drop your email — we'll send a magic link so you can post it. No password.
Not a member of the community? Join today.
Join the community →