Unit economics is the profit calculation for a single unit of your business: one customer, one order, one subscription. You subtract direct costs from the average order value, multiply what is left by the number of purchases a customer makes over their lifetime, and compare the result with what you spent to acquire them. If the difference is negative, every extra sale deepens the loss and ad spend only speeds it up.
Key takeaways
A unit is the smallest thing your business earns or loses money on: a customer, an order, a subscription, a ride, a retail outlet. Unit economics compares two numbers: the profit one unit brings over its lifetime and the cost of acquiring it.
The choice of unit shapes the whole calculation, so fix it before you start. Subscription and service businesses usually take the customer as the unit, retail and delivery take the order, and a marketplace takes the transaction between the two sides.
Mistake Marketplaces and aggregators often run the numbers on gross volume. 16 Startup Metrics by Andreessen Horowitz separates the two explicitly: GMV is the total sales volume transacting through the marketplace, while revenue is only the share the marketplace takes, usually only a fraction of GMV. Run the calculation on revenue.
Five figures are enough for a basic calculation: acquisition cost, average revenue per customer, margin share, purchase frequency and churn. The definitions below come from primary sources, because calculations inside a company diverge not on the arithmetic but on what each department means by these words.
Retention is the other side of churn: the share of customers who stayed. It is only meaningful by cohort, meaning a group of customers acquired in the same month and the share of them still paying in each following month. An average across the whole base hides the fact that newer cohorts may behave worse than older ones.
LTV is the present value of the future net profit from a customer over the duration of the relationship, not the sum of their payments. 16 Startup Metrics calls it a common mistake to estimate LTV as the present value of revenue, or even of gross margin, instead of net profit.
The difference is practical. When revenue goes into LTV, the model looks profitable at any CAC, because an order value is always above zero. Once cost of goods, delivery, payment fees and the cost of serving the customer come out of that order value, some acquisition channels stop paying for themselves.
The working compromise used in SaaS is to calculate LTV on margin. The SaaS Metrics 2.0 definitions write the formula as ARPA multiplied by the gross margin share and divided by the monthly churn rate. It is stricter than a revenue calculation and gentler than a full net profit one, and it is the version used in the rest of this text.
The formulas below sit in one table so you do not have to collect them piece by piece. Each one is calculated for a single segment and a single period.
| Metric | Formula | What goes in |
|---|---|---|
| CAC | Sales and marketing expenses divided by the number of new customers | Advertising, salespeople's salaries, contractors, referral fees, credits and discounts for the chosen period |
| Margin per order | Average order value minus direct costs of the order | Cost of goods, delivery, packaging, payment provider fees |
| LTV on margin | ARPU multiplied by the margin share and divided by the monthly churn rate | For one-off purchases it is simpler: margin per order multiplied by the average number of orders per customer |
| LTV to CAC ratio | LTV divided by CAC | Both figures for the same segment and the same period |
| CAC payback period | CAC divided by ARPU multiplied by the margin share | The result comes out in months |
The CAC, LTV and payback formulas follow the SaaS Metrics 2.0 definitions. A single company-wide CAC hides the fact that one channel pays back three times over while the one next to it never does, so repeat the calculation by channel, by city and by customer type.
Take an online shop with repeat purchases. The figures are illustrative and only show the order of the steps: substitute your own.
Then repeat the same calculation by channel. If paid search costs 800,000 soum per customer and referrals cost 150,000, the average CAC of 500,000 describes neither of them, and no budget decision should rest on it.
The LTV to CAC ratio shows how much cushion the model has, and the payback period shows how much cash has to sit in the business until a customer returns what was spent on them. The benchmarks below come from SaaS practice, and the author of the method states plainly that they are guidelines and there are situations where breaking them makes sense.
These numbers do not transfer to retail, services or delivery, where the cost structure and the purchase frequency are different. It is more useful to compare yourself with yourself: with last quarter and across your own channels.
| What you get | How to read it | What to do |
|---|---|---|
| LTV below CAC | The customer never pays back and more advertising deepens the loss | Stop increasing the budget; raise margin and repeat purchases, or change the channel and the segment |
| LTV slightly above CAC | The model pays back with no cushion: a small error in the churn estimate flips the result | Recalculate by segment, switch off the most expensive channels, cut direct costs per order |
| LTV well above CAC, short payback | The model can take scaling | Increase the budget in the channels where the ratio holds and recalculate it after every increase |
| LTV well above CAC, long payback | The model is profitable but growth runs into a cash gap | Work out how much cash the payback period needs and plan working capital in advance |
The data for unit economics sits in three places: the CRM, the books and the ad accounts. The calculation takes hours when the lead source, the deal amount and repeat purchases are recorded automatically, and weeks when they are reconstructed from chat history and what managers remember.
Connecting ad spend to payments is the job of end-to-end analytics, and we cover how it works in our article on end-to-end analytics from advertising to profit. If the data is not in any system and it is unclear where it goes missing, start with a business audit: it shows which processes leave no trace at all.
Mistakes in unit economics are rarely arithmetic. Almost always they come from badly assembled data that makes the model look better or worse than it is. Walk through this list before you make a budget decision.
Syntra Systems comes at this task from the data side. We start by looking at what the CRM actually records: whether the lead source is there, whether repeat deals are visible, whether the amount in the system matches the amount in the books. Until those fields exist, every calculation is an estimate from memory and the budget is allocated blind.
Then we set up the reports that let the calculation be assembled without manual work: lead sources, stage conversion, money in the pipeline, repeat sales by cohort. This is part of putting sales in order, from $3,000, and the page on CRM systems lists what the work covers. What to do with the resulting numbers is the subject of our article on turning analytics results into an action plan.
Let’s discuss your project
Tell us what you need, and we will estimate the timeline and cost and suggest a solution.
A profit and loss statement shows the result of the whole company for a period, while unit economics shows the result on a single customer or order. A company can be profitable overall and lossmaking in a new channel: profit from older customers covers the loss on new ones, and the summary report never shows it.
A small business needs it more than a large one: the safety margin is thinner, and a mistake in the ad budget hits harder. No complex tools are required — a spreadsheet and discipline in tracking where each customer came from, what they paid, and what the order cost are enough.
Services usually take the customer as the unit and project work takes the project. When projects differ widely in size, calculate by type: small, medium and large. Team time belongs in the direct costs, otherwise the project margin comes out overstated.
At every meaningful change: a new acquisition channel, new prices, a new product, new supplier terms. In a quiet period once a quarter is enough. A calculation from last year may describe a business that no longer exists.
Do not scale. Find the lever first: raise the average order value, increase the share of repeat purchases, cut direct costs or change the channel and the segment. Test each hypothesis with a calculation rather than a feeling, and switch growth back on once the unit turns profitable.
A spreadsheet is enough for the first calculation. A system becomes necessary when the calculation repeats every month and by segment: the data then has to arrive from the CRM and the books on its own, or half the time goes on reconciling figures instead of making decisions.
