Home·Insights·Unit economics: the language investors think in

Unit economics: the language investors think in

Totals persuade amateurs. Professionals ask two questions: what does one unit earn, and what does it cost to acquire.

Project & Investment2026-10-016 min readAtlas Strategy Group

Present a revenue total to a professional investor and you will get the same reflex as showing a magician a card trick: polite interest, no belief. The totals are already known to be survivable fiction — every number in them is a sum. What earns attention is the conversation underneath: what one unit of the business earns, what it costs to acquire, and how long the money takes to come back. That conversation has a name, and it is the working language of every serious diligence process.

Choose the unit honestly

Everything depends on the unit. A customer, an order, a contract, a location, a delivered project — the right unit is the one where the economics actually concentrate: where revenue is earned, cost is incurred and decisions are made. The test is simple: if the unit is healthy but the business is not, or the reverse, you have picked the wrong unit. Subscription businesses live at the customer; transaction businesses at the order; project businesses at the engagement. Companies that skip this step compute impressive ratios over nothing.

The two questions and their follow-ups

The first question: what does the unit contribute — revenue minus the direct, variable costs of serving it, over the unit's honest lifetime with this business. The second: what does it cost to acquire — all spend that exists because this unit was acquired, marketing and sales included. The relationship between the two numbers, and the time it takes the contribution to repay the acquisition cost, is the spine of the business's arithmetic. The follow-ups are where it gets interesting: does contribution improve as cohorts age (real retention) or decay (growth bought by discounts)? Does the newest cohort look like the old ones, or is the machine quietly re-pricing itself downward as it scales?

Read at the unit level, the business stops being a story about growth and becomes a story about replication. Investors fund replication.

What unit-level reading catches that the P&L hides

A profit-and-loss statement can look healthy for a surprisingly long time while the machine underneath is failing — old cohorts still contributing, new ones acquired at a loss, the totals rising on momentum and falling on contact. Unit-level reading catches the failure early, at the cohort where it begins. It also catches the opposite: a business whose totals look modest but whose every cohort pays back faster than the last — the most fundable shape a company can have, and invisible in an annual report.

Where unit arithmetic misleads

  • Early cohorts flatter. The first customers are friendlier, cheaper to serve and often unrepresentative; unit economics computed on them describe a market that is about to disappear.
  • Shared costs hide in averages. Support, infrastructure and management allocated as a flat percentage quietly misstate both the best and worst units.
  • Lifetime is a forecast. Retention assumed for five years earns money that arrives, if at all, in years the reader may never see. The honest version shows contribution at fixed horizons — the money that has actually come back.

Unit economics is also where the financial model gets its drivers — a model built on honest units survives diligence, and a model built on growth percentages does not. It is a standard part of the project and investment practice: before the business is shown to anyone, the business is read at the level where it will be judged.

All insights →