)
)
Professional services billing: summary and key takeaways
The real choice: Professional services billing decides who absorbs estimating error, and now who keeps the hours AI saves.
On a 20% overrun: Uncapped time and materials holds a 50% margin, while fixed fee, milestone and capped hourly billing drop to 40%.
On a 20% AI saving: Fixed fee rises to a 60% margin, while time and materials holds 50% but earns $6,000 less gross profit.
The invoice test: Any model protects margin only when invoices come from approved time, current rates and signed change orders.
Your billing model decides who pays when an estimate turns out wrong, so choose it per engagement rather than by habit. Professional services billing is how a professional services firm turns delivered hours, phases or ongoing access into invoices a client agrees to pay.
Before I joined Teamwork.com, I managed agency accounts where underbidding at the pitch caused most of the scope disputes that followed. That's why this guide treats the model as a risk decision. One $60,000 engagement runs through every model below, so you can see the margin each leaves before you sign.
Four billing models, four different risk owners
Each billing model answers the same question differently: when the hours don't match the estimate, who pays for the gap? The four common models split that risk like this:
Time and materials (T&M) billing: The client pays for approved hours at agreed rates plus expenses, so the client carries overrun risk.
Fixed fee billing: The client pays one agreed price for a defined scope, so your firm carries overrun risk and keeps any savings.
Milestone billing: A fixed fee is invoiced in stages as deliverables are accepted, so risk stays with you and cash arrives in steps.
Retainer billing: The client pays a recurring monthly fee for a set allowance of hours, and your overage terms decide who carries extras.
One variation matters for the worked example. A not-to-exceed cap puts a ceiling on time and materials: the client pays hours up to the cap, and you absorb the rest. Value-based pricing ties the fee to a result, so it sits outside this comparison.
Definitions tell you how each model works, but only numbers tell you which one to sign.
The risk-owner test: one $60,000 engagement, five ways to bill it
The fastest way to choose a model is to run one estimate through every option and see what's left when the hours move. All figures here are illustrative.
Say an engagement is estimated at 400 hours, billed at $150 an hour, with a blended delivery cost of $75 an hour. That gives a planned fee of $60,000, delivery cost of $30,000 and gross profit of $30,000. Gross margin is gross profit as a share of revenue:
At plan, that's 30,000 divided by 60,000, a 50% margin. In the overrun case, delivery takes 480 hours and cost rises to $36,000. In the AI case, the same scope takes 320 hours and cost falls to $24,000.
Two models need a little more setup. Milestone billing splits the fixed fee 30/40/30, so invoices of $18,000, $24,000 and $18,000. The retainer is $15,000 a month for four months, covering 100 hours, and in the overrun case the team works 120 hours a month.
Run both shocks through all six options and one pattern stands out. On an overrun, only uncapped time and materials and a retainer with an overage clause pass the extra hours to the client. Every other model hands the 20% back to you.
Billing model
Every row assumes the same $75 cost on every hour, and the planned case is $60,000 at a 50% margin for all six.
What the overrun does to each model
Uncapped time and materials protects margin on an overrun without any extra clause. The client pays all 480 hours, revenue reaches $72,000 and margin holds at 50%. The trade-off is that a buyer who wants budget certainty will ask for a cap.
A $60,000 not-to-exceed cap makes time and materials behave like a fixed fee on the downside and an hourly rate on the upside. You bill $60,000, absorb the last 80 hours, and margin falls to 40%. In the AI case you bill only 320 hours, so the cap gives you a fixed fee's risk without a fixed fee's upside.
A plain fixed fee takes the same hit on the overrun: gross profit drops to $24,000 and margin to 40%. It earns that back when hours fall, because the fee doesn't move and margin rises to 60%.
Milestone billing has fixed fee margins with a cash-timing twist. The final $18,000 waits on sign-off of the last phase, and in this example the overrun lands in that final phase. So the overrun that costs you margin can also hold back 30% of the fee.
PwC's US guide to measures of progress under ASC 606 says measuring progress by contract milestones is unlikely to be appropriate when significant work happens between them. So a milestone billing schedule doesn't automatically set revenue timing; confirm with your auditor.
The retainer rows show that contract wording matters more than the model name. Without an overage clause, the extra 20 hours a month are free to the client and margin drops to 40%. With overage billed at $150 an hour, the extra hours become revenue:
That lifts revenue to $72,000 and keeps margin at 50%. The retainers I managed agency-side had no overage mechanism, so extra requests inside the fee came straight out of the account's margin. Put the overage rate and rollover rule in the contract before month one, and use the retainer project template to track both.
Which model to choose: rules of thumb from the worked example
These thresholds are rules of thumb drawn from the math above, not research benchmarks, and each one starts from the same number. Measure estimate error on your last five comparable engagements:
In the overrun case, that's 80 divided by 400, or 0.20. With that number in hand, apply these rules:
Choose uncapped time and materials when: scope is genuinely unknown, as in early scoping, and the client accepts a range instead of a ceiling.
Choose capped time and materials when: the client needs a ceiling, and you can set the cap at estimate plus your typical overrun.
Choose fixed fee when: estimate error on comparable work stays under about 15%. Above that, quote fixed fee only with a change-order clause.
Choose milestone billing when: the engagement runs longer than about three months. Keep the final milestone at or below 30% of the fee.
Choose retainer billing when: demand recurs month to month. If actual hours exceed the allowance in two of three months, reprice at renewal.
Three of these rules depend on contract wording as much as the model. A professional services agreement that defines change orders, overage rates and acceptance terms turns them from rules of thumb into terms you can bill.
How AI changes the hourly billing math
On time and materials, every hour AI saves comes off the invoice, so the client keeps the saving unless the contract says otherwise. In the worked example, time and materials still shows a 50% margin at 320 hours. But gross profit lands at $24,000, $6,000 below plan for the same scope.
Fixed fee keeps the full $60,000, so the same saving lifts margin to 60%. At firm scale, the hourly side adds up fast: SPI Research's 2026 Impact of AI on Professional Services report sizes it for a 100-person firm. Every 10% of time-and-materials effort AI removes puts about $1.9mm of revenue at stake, unless the work is repriced or the hours are sold again.
That leaves two ways to keep the saving: reprice repeatable work away from hours, or sell the freed capacity again. Selling hours again only works if you track your utilization rate closely enough to see the gap open.
At Teamwork.com, I've built AI into my own reporting and project admin, and it hands back about a day a week. That day only counts because I can see where it went, and clients will ask the same about hours they once paid for.
Start with the repeatable work where AI cuts the most hours, and move it to a fixed fee priced from what comparable engagements actually cost. Keep time and materials for scoping and advisory work, where hours still track value. The method in how to price professional services projects covers building that fee from actuals.
Before you compare models, add AI tool cost to delivery cost, because the worked example leaves it out. Once you know what a deliverable truly costs, you choose where the gain goes: bank it as margin, or price sharper to win work.
What makes professional services billing match the work
Professional services billing holds margin only when invoices come from the record you price from: approved time, current rates, signed change orders and budget burn. Miss one, and the margin you priced and the margin you invoice drift apart.
Approved time is the base for every model, including fixed fee, because it's how you measure estimate error and margin. A timesheet approved weekly by a project lead gives finance hours it can defend, while one rebuilt at invoice time gives a guess. The guide to tracking billable hours covers the capture habits that make weekly approval quick.
Each approved hour needs two rates: the bill rate from the client's rate card and the cost rate of the person who did the work. Without the cost rate, you see revenue but not margin, and a senior person on junior-rate work stays invisible.
Rates cover the hours you planned, and change orders cover the ones you didn't. A change order is a signed amendment that adds scope, fee or time before the extra work is billed. A change request that never becomes a change order leaks fixed fee and milestone margin, because the hours land in delivery with no fee attached.
Budget burn alerts close the loop. As a rule of thumb, if a fixed fee engagement burns 80% of budget before half the deliverables are done, review scope that week. If your budgets still live in a spreadsheet, the project budget template gives you a phase-level structure to set alerts against.
Accuracy also affects cash. In Atradius's 2026 US payment practices survey, late payments affected an average of 22% of B2B receivables across industries. You can't fix a slow payer, but an invoice backed by approved time and a signed change order removes the easiest grounds for a query.
The five-line pre-invoice check
For layout, payment terms and chasing late payers, how to bill a client covers the mechanics. Getting the numbers right is a separate job, so run these five checks before any invoice leaves:
Time approved: every hour on the invoice has been approved by the project lead for this period.
Rates current: bill rates match the client's rate card and contract, including any agreed increases.
Scope signed: every out-of-scope hour maps to a signed change order or is written off on purpose.
Budget reconciled: amount invoiced, budget burned and budget remaining agree for fixed fee, milestone and retainer engagements.
Terms attached: milestone acceptance, overage hours and payment terms on the invoice match the agreement.
Where this approach stops working
The checklist assumes reliable delivery data and comparable past work, and the advice weakens fast when either is missing. New service lines and small practice areas have too few comparable engagements to measure estimate error, so the 15% rule has nothing to stand on. With no history, I'd start on capped time and materials with a visible buffer, then move to fixed fee once actuals exist.
Poor time data breaks every row of the table. If timesheets are filled in late or from memory, the actual hours are a guess, and so is the margin. Fix capture first, because a weekly approval habit matters more than which model you sign.
Value-based and outcome deals sit outside this math, because the fee follows a result rather than hours or scope. Business Insider reported in November 2025 that a McKinsey managing partner said about a quarter of the firm's global fees come from outcomes-based pricing. That's one large firm's self-reported figure, and McKinsey says the model predates AI.
For a mid-size firm, the lesson is narrower: an outcome fee still needs a cost floor, and only delivery data gives you one.
Billing from delivery data with Teamwork.com
)
Teamwork.com captures the delivery data professional services billing depends on, then passes what to bill to your finance system. Teamwork.com is the agentic PSA that connects projects, resources and financials in one place. Approved time, budgets for each billing model and invoices from logged time sit in one workflow, while QuickBooks or Xero stays your system of record.
Price and cost every hour before it's invoiced: estimate error and margin are guesswork when timesheets arrive late and carry only a bill rate. Time Tracking logs time against tasks as delivery happens, sends timesheets for approval, and applies each person's billable and cost rates. You get actuals for every row of the risk-owner test, ready to measure estimate error on the next quote.
Spot an overrun before the final invoice: fixed fee and retainer overruns often surface only when the last invoice is built. Budgeting & Profitability sets a fixed fee, T&M or retainer budget per engagement, tracks margin against your target and alerts you at a spend threshold. Retainers carries underspend or overspend into the next period, so the overage conversation starts from real numbers.
Invoice from the hours that were approved: rekeying hours into an invoice is where approved time and billed time stop matching. With invoicing and accounting integrations, you build client invoices from logged time and send them to Xero or QuickBooks. Finance keeps receivables, revenue recognition and the ledger where they are, now fed by approved delivery data instead of a reconciliation spreadsheet.
SugarCRM unified projects, time tracking and billing in Teamwork.com, and credited less than $20K on more than $10M in annual invoicing. When the record that prices the work is the same one that bills it, the billing model you chose is the margin you keep.
)
)
)
)
)
)
)