Most unprofitable projects are not badly run. They are badly priced, and the damage is locked in before the kickoff call ever happens. If you want to calculate project profitability with any confidence, you have to do it while the quote is still a draft — when the scope, the hours and the number are all still yours to change. This is a spreadsheet exercise, not an accounting one, and you can build the first version in an afternoon.
What follows is the arithmetic, one worked example, and a decision rule you can apply before you reply to the proposal request.
What Profitability Actually Means on a Single Project
Four numbers describe the economics of one engagement. Everything else is commentary.
- Revenue. What the client actually pays you, after discounts, payment processing fees and any platform cut. Not the headline number on the proposal.
- Direct cost. Delivery hours multiplied by your loaded cost per hour, plus every non-labor cost the project causes: subcontractors, stock assets, licenses, travel, ad spend you front.
- Contribution. Revenue minus direct cost. This is what the project leaves behind to pay for rent, software, admin time and profit.
- Effective hourly rate. Revenue divided by total delivery hours. The single most useful sanity check on a fixed fee.
The word that does the work here is loaded. If you pay a designer $95,000 a year, that person does not cost you $95,000. Add payroll taxes, benefits, equipment and software, and you might be carrying around $119,000. Then divide by delivery hours, not paid hours. A full-time employee is paid for roughly 2,080 hours a year, but sick days, holidays, internal meetings, sales support and admin eat a large share. If you only budget 1,300 genuinely billable delivery hours, your loaded cost is about $91 per hour, not the $46 the payroll number suggests.
Use your own numbers for this. The ratio differs a lot between a solo consultant, a four-person studio and a firm with a project manager on salary.
The Five Inputs You Need Before You Can Calculate Project Profitability
- Loaded cost per delivery hour, per role. One figure for each person or role who will touch the work, including yourself. Founders who price their own time at zero produce permanently misleading estimates.
- An hour estimate broken down by phase and role. Not a single total. Discovery, design, build, review, launch, handover — each with the role attached. Totals hide the phase where you always overrun.
- Non-labor direct costs. Anything you would not spend if this project did not exist.
- A contingency percentage. Based on how this type of work has actually gone for you, not on optimism. Tighter for repeat work with a known client, wider for anything new.
- Overhead absorption per delivery hour. Take your annual fixed costs — rent, software, insurance, accounting, non-billable admin salary — and divide by the delivery hours your team can realistically produce in a year. That gives you the overhead each project hour has to carry.
Five columns. That is the entire model. The difficulty is never the formulas; it is being honest in the cells.
A Worked Example: A Fixed Fee That Looks Fine and Is Not
Assume a three-person studio quotes $18,000 for a ten-week brand and website project. Loaded costs: you at $75 an hour, a senior designer at $91, a contract developer billing you $65.
The estimate by phase:
- Discovery and strategy — you, 20 hours — $1,500
- Design — senior designer, 90 hours — $8,190
- Build — contract developer, 60 hours — $3,900
- Project management, QA and client communication — you, 25 hours — $1,875
That is 195 hours and $15,465 of direct labor. Add non-labor costs: $300 for stock assets and fonts, $120 for staging and hosting during the build, $1,200 for a subcontracted copy pass. Direct cost becomes $17,085.
Contribution is $915 on $18,000 of revenue. Roughly five percent. At this point many people stop, see a positive number, and send the quote.
Now apply the two inputs most estimates skip. A 15% contingency adds about 29 hours and roughly $2,300, pushing direct cost to $19,385 — already above the fee. And the studio’s fixed costs run $9,000 a month, $108,000 a year, against about 3,900 deliverable team hours. That is about $28 of overhead per delivery hour, so 224 hours of work must carry around $6,270.
The real result is an operating loss of roughly $7,650. The effective hourly rate is $80 against a true breakeven of about $115 per hour. To earn a 20% operating margin, this project needed to be priced near $28,000.
A positive contribution number tells you the project pays for itself. It does not tell you the project pays for your business.
The useful part is that this took about fifteen minutes and happened before anyone committed. The options are all still open: cut the design phase, drop the subcontracted copy, raise the fee, or decline and keep 224 hours of capacity free for better work.
A Decision Rule You Can Apply in Ten Minutes
Resist the urge to adopt a generic margin benchmark. Your floor is not a benchmark; it is arithmetic. Run these five checks on every quote above a size that would hurt you to get wrong.
- Effective rate versus true breakeven rate. Revenue divided by estimated hours, compared with loaded labor cost plus overhead per hour. Below breakeven, the answer is no, regardless of how much you like the client.
- Contribution after contingency. Apply the contingency first, then read the margin. A margin that only survives a perfect delivery is not a margin.
- Contingency size as a scope signal. If you need more than about 25% contingency to feel safe, the scope is not defined well enough to fix-price. Quote a paid discovery phase instead.
- Capacity share. If one project consumes more than roughly a third of a quarter’s delivery capacity, price in the concentration risk and check what you are displacing.
- Cash timing. Profit and cash are different problems. If most of the fee arrives after delivery while you are paying subcontractors monthly, fix the payment schedule before you fix the price.
Fail one check, renegotiate scope or price. Fail two, walk away. Writing the rule down before you are emotionally invested in a specific client is most of the value.
Close the Loop: Compare the Estimate With What Happened
An estimate you never check is a guess with formatting. The comparison is what turns pricing from instinct into a skill, and it needs only three numbers per finished project: estimated hours versus actual hours, estimated direct cost versus actual, and planned contribution versus realized contribution.
Keep it light or it will not happen. Any project that overruns hours by more than 20% gets a fifteen-minute review with two outputs: one sentence naming the cause, and one change to how you estimate that phase next time. Common causes repeat — unbilled revision rounds, client response delays that stretch project management hours, a discovery phase that was really design.
After five or six closed projects you will have something far more valuable than a margin target: a contingency percentage and a set of phase estimates grounded in your own history. If you want a structure that already pairs the pre-quote estimate with post-project actuals and ties both to capacity, the Project Profitability & Capacity Planner is built around exactly that loop, and there are more operational walkthroughs like this one on the Cursiqa blog.
Where a Spreadsheet Stops Being Enough
Be honest about the ceiling. A spreadsheet handles project-level economics well, and it handles these situations badly:
- No time tracking. Without recorded actual hours, the variance review is fiction. This is the real dependency, not the tool.
- Shared people across many live projects. Once six or more concurrent projects compete for the same specialists, you need resource scheduling, not a monthly capacity column.
- Beyond roughly eight to ten delivery staff. Per-person utilization forecasting and role-level cost updates become maintenance work nobody owns.
- Retainers and long multi-phase contracts. Revenue spread across periods needs real recognition rules, and that is a conversation with your accountant, not a formula.
- Nobody maintains it. A model only one person updates dies when that person gets busy.
Until you hit one of those, a spreadsheet you actually fill in beats a project system you bought and abandoned.
If you would rather start from a model than a blank sheet, the Cursiqa shop has the Project Profitability & Capacity Planner with the estimate, actuals and capacity views already wired together. Either way, build the quote-stage check first — that is where the decision still belongs to you.