A P&L answers one question: did the company make money last month? Useful — but notice what it can’t answer. Which customers made you the money? Which jobs, products, or locations? If you doubled sales next year, would profit double, or would it disappear into the costs of delivering? Whether growth is worth pursuing at all lives in those answers, and the P&L, which averages everything into totals, is structurally incapable of giving them. A company can grow revenue every year and quietly grow its unprofitable half faster — the totals will look fine right up until they don’t.
Unit economics is the discipline of answering the questions the P&L can’t: taking the business apart into its repeatable unit — a customer, a job, a truck, a location, a project — and asking of each one, what do we earn on this, all-in? It’s the closest thing finance has to taking the engine out and putting each cylinder on the bench.
Picking the unit, and telling yourself the truth about costs
The unit should be the thing your business naturally repeats. For a service firm it’s usually the client or engagement; for contractors, the job; for distribution, the customer or route; for multi-site operators, the location. Choose the level at which decisions actually get made — you can’t act on “average profitability” but you can act on this customer, this route, this branch.
Then comes the step where most homegrown analyses flinch: loading the unit with its honest costs. Revenue per unit is easy. Direct costs — materials, direct labor — usually exist somewhere. The truth-telling starts with everything in between: the delivery cost nobody allocates, the unpaid change orders, the hours your best people spend servicing the demanding account, the warranty work, the payment terms (a customer who pays in 90 days costs real financing money that appears on no job report). None of this needs activity-based-costing perfection — reasonable allocations, honestly applied, beat precise fictions. The standard is simple: if the unit vanished tomorrow, what costs would vanish with it? That’s the unit’s cost, whatever the chart of accounts says.
What the bench test usually finds
Run this analysis on a company that’s never done it and the results tend to follow a consistent pattern: a small share of units generates most of the true profit; a large middle roughly breaks even; and a bottom group — often including some of the biggest, oldest, most beloved accounts — loses money on every transaction, subsidized invisibly by the top. The revenue-ranked customer list and the profit-ranked customer list are different lists, and the gap between them is the single most decision-relevant fact in the company.
Because once the lists exist, decisions that were vague become mechanical. Pricing: the money-losing accounts get repriced with specifics rather than nerve — and the conversation is transformed when you know the number (“we need 8% and 30-day terms to continue”) instead of fearing it. Sales targeting: commissions and effort point at the profitable profile, not the largest. Growth: you now know whether scaling means multiplying the top of the list or the bottom — which is the difference between growth that compounds and growth that merely enlarges the problem. And capacity: when the shop is full, the bottom list is the answer to “full of what?” — pruning break-even work to make room for profitable work raises earnings without adding a single cost.
Your revenue-ranked customer list and your profit-ranked customer list are different lists. Everything useful lives in the gap.
From one-time study to instrument
The first analysis is a project — a few weeks of allocation decisions and spreadsheet work, ideally with someone experienced enough to keep the allocations honest. But the payoff compounds only when it becomes reporting: unit-level margin refreshed monthly or quarterly, sitting alongside the P&L in the owner’s regular package, so drift gets caught while it’s still a conversation instead of a year-end autopsy. This is core FP&A plumbing, and — like the rolling forecast — it’s the kind of instrument a fractional CFO typically installs in an early quarter and your existing team maintains thereafter.
One more audience cares about these numbers, and it’s worth knowing years in advance: buyers. A company that can show margin by customer and cohort isn’t just better managed — it’s provably better managed, which is the difference diligence teams pay for. Unit economics is how you find the profit while you own the business; it’s also how you document the quality of the earnings when you sell it. The starting move is small: pick your unit, take your ten largest, load them honestly, and rank them by true margin. The list usually contains at least one surprise.
Chief Perspective builds unit-level profitability analysis as part of its FP&A practice with middle-market companies. If you’ve never seen your customer list ranked by actual profit, let’s talk.
Common questions
How precise do the allocations need to be?
Directionally honest beats decimally precise. The decisions — reprice, prune, pursue — turn on which third of the list a unit sits in, not on the second decimal of its margin. Refine allocations over time; don’t let perfect costing delay the first ranking.
What if I find a big customer is unprofitable — I can’t just fire them.
Rarely should you, immediately. The sequence is: reprice with specifics, restructure terms, reduce the cost to serve — and only then, if nothing moves, plan a graceful exit as capacity fills with better work. The point of the number is that it turns “we can’t afford to lose them” into an actual calculation.
Is this different from gross margin on my P&L?
Yes, in the two ways that matter: it’s per-unit rather than averaged, and it’s loaded with the real costs to serve that gross margin ignores. In our experience, a healthy average gross margin hiding a money-losing bottom quartile is common rather than unusual.