Unit Economics: Why Revenue Growth Won’t Fix Your Model

If I could force every SME owner to answer one question before they scale, it would be this: what does it actually cost you, fully loaded, to deliver one unit of what you sell — and how does that compare to what you get paid for it? That single question is what unit economics answers.

Not the blended, company-wide gross margin. Rather, the economics of one patient visit, one delivery, one project, one customer, one MW installed. Because a business can grow revenue for years while its fundamental unit economics sit upside down, and the growth itself is what eventually kills it — faster, in fact, than standing still would have.

The trap: growth accelerates a broken model

If you lose money on every unit you sell, selling more units doesn’t fix the problem. Instead, it accelerates the losses. This isn’t a fintech-specific issue either, though it’s most famous there. I see the same structural error across every industry I work in.

Transaction-level units: fintech, logistics, healthcare

Fintech. A lending or payments platform calculates its cost to acquire a customer (marketing spend, onboarding, KYC/compliance cost) and compares it to the lifetime value of that customer (net revenue over the relationship, adjusted for churn and credit losses). When CAC exceeds LTV, or the payback period stretches beyond what the business can fund, every new customer makes the cash position worse rather than better — even as revenue and user count climb.

Logistics. Here the relevant unit is cost-per-delivery or cost-per-route: fuel, driver time, vehicle depreciation, and overhead allocated per delivery, against what the client pays per delivery. I’ve seen last-mile operators win large contracts at rates that don’t cover fully-loaded cost-per-drop once you allocate fuel and driver overtime properly — a “big win” that quietly drains cash with every delivery made.

Healthcare. Similarly, the unit is cost-per-visit or cost-per-procedure: clinician time, consumables, facility overhead, and — critically — the actual reimbursement rate net of any insurer discount or delay-related cost of capital. A clinic can look busy and still run procedures at a contribution margin close to zero if the payer mix has shifted toward lower-reimbursement plans.

Project-level units: construction, real estate, energy

Construction. The unit is the project: fully-loaded project cost — direct labor, materials, subcontractors, site overhead, and an honest allocation of head-office cost — against the contract value. Contractors chasing backlog sometimes bid projects at a contribution margin that doesn’t cover the true cost-to-serve once site overhead and delay risk go in properly. Consequently they discover the truth only at project close-out, months too late to act.

Real estate development. The unit is the residential or commercial unit sold: land cost, construction cost, financing cost, and sales/marketing cost per unit, against the achieved sale price per unit — not the launch price, the actual achieved price after negotiation and incentives. Margin compression here stays invisible until someone builds a portfolio-level view by launch cohort.

Energy. Likewise, the unit is often cost-per-MW installed or cost-per-unit of output, with long project cycles that hide cost overruns until long after the business has committed. A services contractor can win volume in a growing sector while quietly eroding margin per project, because competition compresses pricing faster than efficiency gains offset it.

Whatever the sector, every business needs to define its unit, measure it, and watch it move.

Why blended margins hide this

Company-wide gross margin is an average, and averages hide exactly the problem you need to see. A business can post a healthy 25% blended gross margin while selling 30% of its units at negative contribution margin, subsidised by a smaller set of highly profitable ones. Therefore any growth strategy built on the blended number will, by default, grow the loss-making segment right alongside the profitable one — because nothing in the reporting distinguishes them. This is the same blind spot that segment-level EBITDA analysis exists to close.

Building a real unit economics model

The model doesn’t need to be complicated to be useful. At minimum:

1. Define the unit clearly — the smallest repeatable thing you sell (a delivery, a visit, a project, a customer, a unit sold).
2. Fully load the direct cost — not just the obvious materials and labor, but the overhead that scales with volume (site supervision, dispatch, scheduling, quality control).
3. Calculate contribution margin per unit — revenue per unit minus fully loaded direct cost. This is the number that tells you whether volume helps or hurts.
4. Segment it — by client, by project type, by service line, by region, by payer or channel. The company-wide number is the least useful cut you can look at.
5. Stress-test before committing to growth — before you sign a large new contract, launch a new project, or enter a new market, model the unit economics of that specific decision rather than assuming it behaves like the rest of the portfolio.

None of this requires expensive systems. Standard financial modeling and a disciplined monthly review will get you most of the way.

The practical takeaway

Revenue growth is neither inherently good nor bad. Instead, it amplifies whatever is already true about your unit economics. If yours are sound, growth compounds value quickly. If they’re broken, growth compounds losses just as quickly, and usually with a lag long enough that the business is deep into the problem before anyone notices — the same dynamic that lets overhead quietly outgrow revenue. So before the next growth push, whether that’s a new contract, a new location, or a new client segment, build the unit economics model first. It’s a few hours of work that can save months of chasing revenue that was never worth having.

For a complimentary 30-minute consultation, book a free session with Dixie Business Consulting.

Why Choose Dixie Business Consulting?