Unit economics for a services business: what one billable hour really earns

We've pulled apart a D2C order and a SaaS subscription. Both sold a thing — a box, a seat. A services business sells something stranger: hours of its people's time. You bill $150 an hour, you pay the person doing the work a fraction of that, and yet a lot of agencies, consultancies, and dev shops end the year wondering where the money went. Part three of the series is about where it goes — and it hides in a place the P&L will never point to.

The culprit isn't your rate and it usually isn't your salaries. It's the hours you didn't bill. A person costs you the same whether they sell 50% of their week or 85% of it — and every unsold hour quietly reloads onto the ones you did sell. Get that number right and a modest rate prints money; get it wrong and a premium rate still loses. Same as before, three rungs tell the story, and there's one cost a box and a subscription never have to eat.

What is the "unit" in a services business?

For D2C the unit was one order; for SaaS, one customer across their life. For a firm that sells time, the smallest repeating thing you sell is one billable hour. Not one project (too big and too custom to compare), not one client — one hour of the work, measured for what it truly earns after everything standing between the rate card and your bank account has taken its cut.

And the number you're chasing isn't your billing rate. The rate is a sticker price; the honest figure is what an hour contributes once you account for the discounts you gave, the time you worked but never billed, the salary of the person doing it, and the firm humming in the background to make the work sellable. Three questions get you there: what do you actually collect for an hour, what does the hour cost to deliver, and what does it cost to keep the lights on around it.

The hour that earns less than the rate card says

Let's run one hour all the way down. Our example is a creative agency billing $150 an hour. That's the rack rate — the number on the proposal. But almost nobody collects their rack rate: there's the scope you ran over and ate, the "we'll knock 10% off" on the invoice, the junior's hours you wrote down before the client saw them. That leak has a name — realization — and at a healthy-looking 90% it already quietly clips $15 off every hour before we've counted a single cost.

A hundred and fifty on the rate card; thirty-five in real life. And notice the biggest single deduction wasn't the salary line you'd expect — it was delivery labor at 65% utilization. That $66.57 isn't what you pay the person per hour; it's their fully-loaded annual cost spread across only the hours they actually billed. Bill fewer hours and that number climbs — which is exactly the trap, and it deserves its own look.

Why utilization is the whole game

Here's the mechanism. Your agency biller earns $75,000, and with payroll taxes and benefits they cost you about $90,000 fully loaded. That number is fixed — it's the same in a busy month and a slow one. What isn't fixed is how many billable hours you spread it across.

• At 65% utilization, they bill about 2,080 × 65% ≈ 1,352 hours, so their loaded cost lands at $90,000 ÷ 1,352 ≈ $66.57 per billable hour.
• Push utilization to 80% and they bill ≈ 1,664 hours — the very same $90,000 now costs just ≈ $54 an hour.
• Let it slip to 50% and the cost balloons to ≈ $87 an hour — nearly your entire realized rate, gone before overhead.

You didn't change a single salary, and the cost per hour swung by more than $30. Utilization sits in the denominator of every services cost the way churn sits in the denominator of LTV — a small move in it swings everything downstream. It is the single most important number in a services business, and it's the one that never appears on a financial statement.

R1, R2, R3: the services ladder

D2C operators count in CM1/CM2/CM3; SaaS climbs L1/L2/L3. A services firm has the same three rungs — the realized rate, then the margin after the person, then the margin after the firm. Call them R1, R2, R3.

• R1 — realized rate. Rack rate minus the realization leak. $150 × 90% = $135. The cash you actually collect for an hour of work — always below the number on the proposal, and the first place margin quietly walks out the door.
• R2 — gross margin per hour. Realized rate minus the loaded cost of the person's time, at your real utilization. $135 − $66.57 ≈ $68. The delivery profit an hour throws off — and the rung utilization makes or breaks.
• R3 — net margin per hour. Gross margin minus the firm overhead behind every biller — rent, admin, tools, the account managers and partners who don't bill. $68 − $33 ≈ $35. What an hour is truly worth to the business. Unlike a D2C first order, it's positive here — but a lot thinner than $150 ever suggested.

Read those left to right and the agency works: it collects a real rate, keeps real money after delivery, and still has margin left after the firm. But zoom out to the whole year and a warning light comes on. That biller generates 1,352 hours × $135 ≈ $182,500 of revenue on a $75,000 salary — a 2.4× revenue-to-salary multiple. The old agency rule of thumb is 3× (a third for salary, a third for overhead, a third for profit), and this firm is coming up short. Not because the rate is wrong — because the utilization is.

Build your own hour

Your firm isn't this one. A consultancy bills triple the rate but carries triple the overhead; a lean dev shop lives on 80% utilization; a solo operator has almost no overhead but can only sell a slice of a week they also have to run. Put in your real rate, realization, utilization, and the loaded cost behind a biller, and watch the ladder — and the two reads that decide the whole thing — resolve:

Prefer a spreadsheet? Open the free Unit Economics calculator to run all four business models side by side and download a branded, formula-driven Excel of your own billable hour.

Two knobs move everything. Drag utilization from 65% to 80% and watch net margin per hour jump by half — you added no clients and raised no rates, you just sold more of a week you were already paying for. Then nudge realization down from 90% to 80% and watch a "we'll just eat that scope" habit erase profit that never showed up as a discount anywhere in your books.

The two reads a services firm lives by

Margin per hour is the ladder; these two numbers are what you tape to the monitor. The first is break-even utilization — the share of a biller's week you have to sell just to cover their loaded salary and their slice of overhead. For our agency that's ($90k + $45k) ÷ (2,080 × $135) ≈ 48%. Every hour billed above 48% is profit; every hour below it, you're paying someone to sit on the bench.

The second is the revenue multiple — revenue per biller divided by their salary. It's the services cousin of LTV:CAC: a single ratio that tells you whether the model has room in it. Land near 3× and the classic thirds hold. Come in at 2.4× and something has to give — and now you know exactly which levers move it, because you can see the machinery underneath the ratio instead of just the number.

The cost a box never has to eat: the unsold hour

Go back to the calculator and drop utilization, then watch the year's revenue fall while every cost stays put. This is the thing that makes services different from anything physical: capacity is perishable. A D2C brand's unsold inventory sits on a shelf and sells next month. A SaaS seat you didn't fill today you can still fill tomorrow. But an hour of your team's time that goes unbilled on a Tuesday afternoon is gone forever — you can't inventory it, you can't discount it later, you simply never get it back.

That's why utilization is the lever with no equivalent in the other two models. A subscription can lean on negative churn to grow while it sleeps; a services firm has the opposite physics — it shrinks while it sleeps, one unsold hour at a time. Your entire inventory expires at the end of every day and gets manufactured fresh the next morning. Selling it before it spoils is the business.

Four levers that move the hour

When R2 is thin or the revenue multiple is short, there are only a handful of places to push — and, as always, the one everyone reaches for first (raise the rate) is rarely the fastest.

• Lift utilization before anything else. It's the highest-leverage number on the page because it's in the denominator of every cost. Moving a team from 60% to 70% billable does more for profit than a rate increase your clients will actually fight — and costs you nothing but better scheduling and less bench.
• Plug the realization leak. Scope creep you absorb, discounts you give reflexively, and hours you write down are pure margin with nothing to show for it. Tighter scopes, change orders that actually go out, and a little spine on the invoice defend a rate you already earned — no selling required.
• Then raise the rate — selectively. Rate increases drop almost entirely to R3, but they meet the most resistance, so aim them where you have leverage: new clients, senior work, outcomes that are clearly worth it. A rack rate nobody realizes isn't a rate, it's a fiction.
• Watch overhead per biller as you grow. R1 and R2 can look great while R3 quietly bleeds if the non-billable side — managers, admin, tools, real estate — grows faster than the billing team. The revenue multiple is your early warning: if it drifts below 3×, overhead is usually the reason.