Unit economics asks a simple question: does each additional customer make you money, and how long does it take? A business with poor unit economics gets worse as it grows, which is why the question matters long before profitability does.
The four numbers
Gross margin. Revenue minus the direct cost of delivering the service, as a percentage. For software this includes hosting, third-party APIs, payment processing and the support cost attributable to serving customers. Software businesses commonly run 70–85%; anything materially below suggests a services component or an infrastructure problem.
Customer acquisition cost. Fully loaded sales and marketing spend divided by new customers acquired.
Lifetime value. Gross profit per customer over their expected lifetime. Not revenue — gross profit, which means after the margin above.
Payback period. Months of gross profit required to recover acquisition cost. The number that determines how fast you can afford to grow. See CAC payback period.
Calculating lifetime value honestly
LTV is the number most often inflated, usually through optimistic assumptions rather than deliberate distortion.
The basic form is average revenue per customer per month, multiplied by gross margin, divided by monthly churn rate. A customer paying $500 at 80% margin with 3% monthly churn gives $500 × 0.80 ÷ 0.03 = $13,333.
Three adjustments make it more honest. Use actual observed churn for a comparable cohort, not an aspiration. Apply a discount rate for long lifetimes, since money arriving in year five is worth less than money arriving now. And cap the horizon at three to five years, because a calculation assuming a twelve-year customer life is not a measurement.
The LTV to CAC ratio, and its limits
The widely used benchmark is 3:1 or better. It is a reasonable rule of thumb and it hides two things.
It says nothing about timing. A 5:1 ratio with a 30-month payback consumes enormous cash, and cash is what kills companies rather than ratios.
And it depends entirely on the churn assumption. At 2% monthly churn the ratio looks excellent; at 5% the same business looks marginal. Early-stage companies have too little data to know which is true, which is why payback period — which requires no lifetime assumption — is the more reliable early measure.
Where the calculation goes wrong
Revenue instead of gross profit. Using revenue in LTV overstates it by the entire cost of delivery.
Marketing spend only. Excluding sales salaries, commission and tooling understates CAC substantially.
Founder time at zero. Early-stage CAC looks wonderful when the founder is selling and costing nothing. It triples when the first rep arrives.
Blending segments. A blended figure across SMB and enterprise describes neither. Calculate separately wherever the segments behave differently.
Ignoring expansion and contraction. Both change lifetime value materially, in opposite directions.
Too little data. With six months of history you cannot know lifetime. Say so rather than extrapolating.
Reading the result
Four diagnostic patterns.
Good margin, high CAC, low churn. A viable enterprise business that needs capital to fund long payback. The constraint is cash, not the model.
Good margin, low CAC, high churn. A leaky bucket. Acquisition works and retention does not. See reducing churn.
Low margin, any CAC. A delivery cost problem. Either the infrastructure is inefficient or the business is more services than software, and the usual software benchmarks do not apply.
Everything looks fine but cash is falling. Usually a payback timing problem: the economics work eventually and the growth rate outruns the cash.
Improving them
In rough order of speed.
- Raise prices. Improves margin, lifetime value and payback simultaneously, with no change in cost. The most underused lever at every stage.
- Reduce involuntary churn. Payment recovery is a fortnight of work and improves retention immediately.
- Improve gross margin. Infrastructure efficiency and support automation. See cloud cost optimization.
- Shift the channel mix toward lower-cost acquisition, which takes quarters rather than weeks.
- Tighten the segment. Selling to better-fit customers improves win rate, retention and expansion at once.
When to start measuring
From the first paying customers, accepting that early figures are unreliable. The value is in building the habit and the data collection, so that when you have enough customers for the numbers to mean something, the history exists.
Review quarterly, by segment, alongside the cash position. Monthly is too noisy for most early-stage companies and invites over-reaction to a single large deal.
Unit economics by segment
A blended figure describes an average customer who does not exist. The useful version splits the base.
Calculate the four numbers separately for each segment you serve — by size, by channel, by plan, by industry if it matters. The differences are usually large enough to change decisions: it is common to find one segment with a six-month payback and another with a thirty-month one, both hidden inside a blended fourteen.
Three things follow from that view. The good segment deserves more acquisition spend than it is getting. The poor segment needs either a price increase, a cheaper acquisition route, or to be stopped. And the sales compensation plan should reflect the difference, or reps will continue closing whatever is easiest.
Presenting them to investors
Unit economics appear in every fundraising conversation, and the way they are presented affects credibility more than the numbers themselves.
State your definitions explicitly — what is in CAC, which margin, what churn assumption, over what horizon. Investors see inflated figures constantly, and the fastest way to be trusted is to show the working.
Show the trend rather than a point. Improving economics over four quarters is a stronger signal than a good number in one.
And acknowledge what you do not know. A company eighteen months old cannot have measured five-year lifetime value, and saying so is more convincing than a confident extrapolation anyone experienced will discount anyway.
A simple monthly review
Unit economics only influence decisions if they are looked at on a schedule, and a short recurring review is enough.
Once a month, calculate gross margin from the previous month’s actuals, CAC from the previous quarter’s spend and new customers, and payback from the two. Record them in the same place each time so the trend is visible.
Then ask one question: did anything we did last month improve or worsen these, and do we know which? That question is what turns the numbers from reporting into management. A company that can name the change responsible for a shift in payback is learning; one that watches the number move without explanation is not.
Benchmarks, with a caveat
Comparing your figures to published benchmarks is tempting and should be done carefully, because the definitions behind most published numbers are not stated.
Broadly, software businesses commonly report gross margins in the seventies and eighties, payback under twelve months is the usual venture expectation, and an LTV to CAC ratio of three is the standard threshold. Each of these varies substantially by segment and by how the inputs were calculated.
The more useful comparison is your own trend. Margin improving two points a quarter, payback shortening from eighteen months to fourteen, churn falling — these tell you whether the business is getting better, which is the question benchmarks cannot answer.
Before the numbers mean anything
Unit economics calculated across fifteen customers acquired by the founder describe a situation that will not persist. Acquisition cost is understated, the sample is small, and the customers were probably chosen rather than typical.
That is not an argument for ignoring them, but for understanding where they come from. The period of founder-led sales is where the inputs are discovered rather than measured, and the figures should be treated as a hypothesis until a repeatable process exists to test them against.
Frequently asked questions
What is a good LTV to CAC ratio?
Three to one or better is the usual benchmark. Much higher can indicate underinvestment in growth rather than excellence.
How early can unit economics be measured?
From the first customers, with the caveat that lifetime is unknown. Payback period is measurable much earlier than lifetime value and is more reliable.
Should founder time be included in CAC?
For an honest view, yes, at a notional market salary. Otherwise the figure flatters you badly until the first hire.
Do unit economics matter if we are growing fast?
More, not less. Poor unit economics scale with the business, and growth accelerates the cash consumption rather than solving it.
