Unit Economics
Whether a single customer or transaction makes money once you count what it costs to acquire and serve - the health check beneath any growth story.
What it is
Unit economics is the discipline of asking whether the fundamental repeatable unit of a business - usually one customer, sometimes one transaction or one location - actually makes money on its own. It strips away the headline growth and total revenue and looks at the single unit: what it earns over its life versus what it costs to acquire and to serve. For a subscription business the classic form compares the lifetime value of a customer to the cost of acquiring that customer. If each unit loses money, scaling the business just loses money faster; if each unit is solidly profitable, growth compounds. It is the reality check that separates a healthy business from a subsidised one that looks impressive only because it is spending to grow.
When to use it
Use it whenever a company is growing fast, burning cash, or asking whether to scale - especially subscription, marketplace, delivery, and digital businesses where acquisition spend can mask a broken model. It is the right lens behind 'they're growing 200% a year but losing money - is that fine or fatal?'. It also sharpens pricing and channel decisions. It is less central for stable, mature businesses where the model is long proven, or for one-off strategic questions with no repeatable unit. Do not confuse total-company profitability with unit economics - a company can be unprofitable overall while each unit is healthy (fixed costs not yet spread) or profitable overall while cross-subsidising loss-making units.
How to apply it
Define the unit precisely, then build both sides of it. On the value side: revenue per unit over its lifetime, net of the variable cost to serve it - for a repeat customer, that means contribution per period times the expected number of periods they stay. On the cost side: the fully loaded cost to acquire that unit. Compare them: a durable business needs lifetime value comfortably above acquisition cost, with enough headroom to also cover fixed costs and earn a return. Then interrogate the assumptions that move the answer most - retention/churn (how long the unit lasts) and acquisition cost (which often rises as you scale past the easy customers). Check the payback period too: how long until a unit repays what it cost to acquire, because a long payback strains cash even when the lifetime math works.
Common mistakes
The dominant error is mistaking growth for health - celebrating rising revenue while each new customer loses money, so faster growth means faster losses. Another is being too optimistic on retention: assuming customers stay for years inflates lifetime value and hides a broken unit. People also understate acquisition cost by counting only ad spend and ignoring sales, onboarding, and incentives, and they forget acquisition cost tends to rise as the cheap customers run out. And they use revenue rather than contribution on the value side, ignoring the cost to serve.
Example
A meal-delivery startup is growing users 15% a month and raising money on that curve. Checking unit economics: an average customer generates about 12 of contribution per month (order margin minus delivery and packaging), stays roughly 8 months, so lifetime value is about 96. But acquiring that customer - discounts, referral bonuses, and ads - costs about 130. Each customer loses roughly 34 over their life, so every month of 'growth' deepens the hole, and the impressive curve is actually a cash incinerator. Worse, acquisition cost is creeping up as the easy customers sign up. The unit-economics lens reveals that the priority is not more growth but fixing retention or acquisition cost until the unit turns positive.
Practice it
Practice on growth, startup, and profitability cases by building both sides of the unit - lifetime contribution versus fully loaded acquisition cost - and stress-testing retention and acquisition-cost assumptions. The skill to build is refusing to let top-line growth substitute for per-unit profitability, and knowing which assumption the answer really hinges on.