)
Summary and key takeaways
Delivery drift comes first: Light estimates and unbilled changes drain a fixed fee, because a budget built on bill rates or total hours can't show them as they happen.
Cost rates, not bill rates: Price every budgeted hour at what it costs your firm, or the budget hides the margin it's meant to protect.
Named-risk contingency: Size contingency from specific risks you can cost, and send scope changes to priced change orders instead.
A floor before kickoff: Set a margin floor and the maximum cost the fee can absorb before anyone logs an hour.
Burn next to progress: Compare budget used with work complete every week, and act when the gap passes a threshold you agreed in advance.
Project budgeting for professional services starts after the client has already fixed the fee. Your job is to turn that fee into a delivery budget built from role-level cost rates. Then you set a margin floor before work starts and track burn against progress every week. Done that way, a light estimate or an unbilled change shows up in week three, not at close-out.
Before I joined Teamwork.com, I ran project budgets of roughly $50k to $100k+ on a basic time tracker and spreadsheets. Scope creep and over-servicing did more damage to those budgets than anything else. My lesson: a budget only protects you if it shows an extra request's cost in the week someone agrees to it.
The fee comes first, so the budget does a different job
That extra request lands on a number the client agreed before your budget existed. Most budgeting advice assumes the opposite order: build a budget, then ask someone to fund it. In professional services the fee is settled first, so the budget can't be a request for money. Its job is to protect the gap between what delivery costs you and what you've sold.
A project budget, in the services sense, is the plan for how much delivery cost a sold fee can carry. A cost rate is what one hour of a person's time costs your firm, including salary, benefits and overhead. A bill rate is what you charge the client for that hour.
Two habits hide the margin inside that gap. Budgets built on bill rates or blended hours make every hour look equally profitable, so a senior person on junior tasks never looks like a cost problem. Change requests get absorbed as small favors that draw down hours nobody priced.
In a 2024 BCG survey of C-suite executives, nearly half said more than 30% of their technology development projects were over budget and late. That's a cross-industry survey of tech projects, not services firms. The point still holds: when the fee can't stretch, the budget has to flag an overrun early enough to act.
Project budgeting for professional services in six steps
An early warning is only as good as the budget underneath it. To budget a professional services project, start from the sold fee and its assumptions, then estimate hours bottom-up by role. Price those hours at cost rates and add contingency for named risks. Set a margin floor and the maximum cost the fee can absorb, then baseline the budget and track burn against progress weekly.
One illustrative project runs through every step: a $60,000 fixed-fee website rebuild covering strategy, design and build. The hours and rates are assumptions chosen to keep the arithmetic clear, not benchmarks.
1. Start from the sold scope and fee, and write down the assumptions
Your budget inherits every promise in the statement of work, so record those promises before you estimate an hour. Note the revision rounds the fee covers, what's excluded, the rates you assumed and the billing model. Those notes become the test for every later request: inside the assumptions, it's in the budget; outside them, it's a change order.
When a firm underbids at the pitch to win, the scope creep that follows is a scoping problem, not a client problem. No delivery budget can rescue a fee that was too low from the start. Settle that first with a method for pricing professional services projects, then build the budget on top.
2. Break the work down and estimate bottom-up by role
Bottom-up estimating builds the total from the smallest units of work: you estimate each task's hours, then roll them up. Start with a work breakdown structure listing deliverables, the tasks under each, and the role doing each task. Estimating by role matters because a strategy lead's hour and a developer's hour cost your firm different amounts.
Ask the people doing the work to estimate their own tasks, then check those numbers against what similar past projects took. Top-down estimates still have a place in early quotes, and this guide to project cost estimation methods compares the main approaches.
3. Price the hours at cost rates, and budget the non-billable time
Multiply each role's hours by its cost rate, not its bill rate, because the budget measures what delivery costs you. The fee is your revenue line, and mixing the two makes margin impossible to read. Then add the delivery time nobody invoices: project management, internal reviews and client reporting.
On the accounts I ran, client reporting was manual, and it ate hours every week that never showed up as billable. I'd give reporting its own budget line from day one, sized by how often the client expects an update.
Delivery cost build: $60,000 fixed-fee website rebuild (illustrative hours and rates)
Cost line
The project manager's 50 hours never appear on an invoice, yet they're 11% of the labor cost. Leave them out and the budget looks healthier than the project is.
A cost build is only as reliable as the rate and time data behind it. SPI Research's 2026 Impact of AI on Professional Services report puts numbers on that. Firms with highly ready data report project margins of 43.6% vs. 35.2% for firms whose data is not ready or fragmented. For your budget, that means cost rates and logged hours belong in one connected place, or the planned margin is a guess.
4. Size contingency from named risks, not a flat percentage
Contingency covers the uncertainty in your own estimate, so size it from risks you can name and cost. The GAO Cost Estimating and Assessment Guide states the principle plainly. "A risk analysis should be used to determine a program's contingency funding." GAO wrote that for federal programs, but it transfers to a fixed fee. List each risk, estimate the hours it would add, and price those hours at cost.
In the example, the CMS integration is unproven and could need 30 more developer hours, or $1,950 at $65. The client also has several stakeholders signing off each design, which could add an approval round of 20 design hours, or $1,100 at $55. Together that's $3,050 of contingency, taking the total cost budget to $31,950.
Keep scope changes out of contingency entirely. A request outside the step 1 assumptions becomes a priced change order, not a draw on the reserve. Once contingency funds new scope, it can't cover the estimate risk it was sized for.
5. Set the margin floor and the trigger point
A margin floor is the lowest project margin you'll accept before you change the plan or the scope. Agree it with the head of delivery before kickoff. Then convert it into a trigger point, the maximum cost the fee can absorb while staying at the floor. With a $60,000 fee and a $31,950 cost budget, the planned margin works out as follows:
That's a planned margin of about 47%. With a 40% floor, the trigger point is 60% of the fee:
The fee can absorb up to $36,000 of cost, leaving $4,050 of headroom above the $31,950 budget. Headroom and contingency do different jobs. Contingency is planned spend for named risks, while headroom is the margin you can lose before the project drops below the floor.
6. Baseline the budget and track burn against progress weekly
Once the client signs off, freeze the cost budget as your baseline, so every later number has a fixed point of comparison. Log time weekly against the same roles and task lists you estimated. Record approved change orders as separate budget lines, so the baseline stays clean and the extra fee stays visible.
This guide to project cost tracking walks through capturing actuals. A budget vs actual tracker template gives you a ready-made layout for the weekly comparison. Burn on its own, though, only tells you how much you've spent.
Reading burn against progress: the week-six check
Sixty percent of the budget gone sounds alarming, but it means nothing until you know how much work is done. Budget burn is the share of the cost budget used so far, and it only means something next to the share of work complete.
At week six of the website rebuild, the team judges the work to be 50% complete. Earned value, the budgeted cost of the work actually done, is 0.5 × $31,950, or $15,975. Actual cost so far is $19,200, which is 60% of the cost budget.
Cost variance is earned value minus actual cost, and a negative result means the completed work cost more than planned. That's how the Department of Energy's earned value tutorial defines it:
If the remaining work runs at the same efficiency, the forecast cost at completion is actual cost divided by the share of work complete:
That forecast lands $2,400 past the $36,000 trigger point, and the forecast margin falls to 36%, below the 40% floor. Measure percent complete from finished tasks or deliverables, never from hours logged, or you're comparing burn with itself.
I'd act at three thresholds, measured as the gap in percentage points between burn and progress. They're my judgment rather than research, so tighten them on long fixed-fee projects and loosen them on short ones.
Budget-control rules (burn vs progress)
Burn minus progress gap
The website rebuild sits at a 10-point gap, which on its own calls for a corrective plan. Its forecast margin is already under the floor, though, so it moves straight to re-planning or a change order conversation. When the project closes, a project profitability variance analysis shows which estimate or decision moved the margin.
When the billing model or project length changes the rules
Put the same week-six numbers on a time-and-materials contract and the overrun lands on the client's invoice instead of your margin. The billing model decides who carries an overrun, so it changes what the budget is for.
On a fixed fee, your firm absorbs every hour past the trigger point, so the budget works as a margin limit. When budgeting projects on time and materials, the client pays for overruns up to any agreed cap. The budget becomes a forecast you share early, so the cap never surprises anyone.
Retainers reset each period, so the risk shifts from one large overrun to small ones that repeat. Run the burn check per period, and treat any period that runs past the retainer's hours as over-servicing to raise with the client. Professional services budgeting on a retainer is mostly about stopping one month's favor from becoming next month's baseline.
Very short projects, under about two weeks, don't need weekly control. Run a midpoint check and a close-out review instead, because a weekly threshold on a project that short mostly fires on noise. If a project is already past its trigger point, start with these options for recovering a project that's already over budget.
None of this works if time isn't logged weekly, because burn built from late timesheets lags reality by exactly the weeks you needed. For ready-made layouts, this project budget template roundup compares the options.
Pro tip: Set one weekly cutoff for logging time and run the burn check the next working day, so your numbers stay current.
How Teamwork.com carries the budget from quote to the week-six check
)
Teamwork.com is the agentic PSA, a professional services automation platform that connects projects, resources and financials. Rates and quotes become the project budget, the margin floor becomes a tracked target, and logged time feeds burn against progress. Project tools track the work but can't manage the money, and traditional PSAs manage the money but teams resist using them. Teamwork.com does both, so the budget runs on real numbers from live delivery, not a spreadsheet you rebuild each week.
Turn the sold fee into a budget without retyping it. Re-entering a quote into a project plan is where rates drift and assumptions get lost. Quoting & Costing Work stores cost and billable rates for every team member, with billable rates set by role, project or client. An accepted quote converts into a project with the quote details carried over, so the hours and cost rates from steps 1 and 3 arrive in delivery intact.
Know where the margin floor sits before the first hour is logged. Without a set floor, nobody can tell when a project has spent its headroom. Budgeting & Profitability runs fixed fee, time and materials, or retainer budgets, split across task lists. You set a profit margin target from step 5, and budget notifications fire when spend reaches your chosen threshold. Time tracking captures billable and non-billable hours, so the coordination time from step 3 counts against the budget.
On one client project, The Brand Leader used budgeting in Teamwork.com to track the time spent in a brainstorm session and the meetings that followed. The team saw it had an hour left before the work stopped being profitable, a visibility its president ties to the wider result:
"By understanding our cost centers and the cost per person per project, we're now more profitable and more efficient." — Kyle Duford, President and Executive Creative Director, The Brand Leader
Spot burn running ahead of progress across the portfolio. One project's week-six problem is easy to miss when every project reports separately. The Project Health Report and Planned vs. Actual Report show task completion next to budget used for every active project. A project with a quarter of its tasks done and over half its budget gone stands out without anyone building a report.
)
Each step keeps the same number in view: the gap between what delivery costs and what the client agreed to pay.
)
)
)
)
)
)
)
)