How to price professional services projects: the four-gate method

I have spent enough time in agencies to know how to price professional services projects badly. The reason most projects lose money has little to do with the pricing model on the proposal. It comes down to two things nobody systematises: what the work truly costs to deliver, and what happens to that cost after the client signs. This guide fixes both with a repeatable four-gate method.

Why pricing a project got harder than picking a model

In my agency years, the pricing conversation was never really about the model. Teams argued over hourly versus fixed fee while the leak sat somewhere else. Most advice treats pricing as a menu problem, as if picking hourly, fixed-fee, retainer, or value-based makes margin follow, but the model is only a container. What goes into it, your true cost and your ability to hold scope, decides whether it holds profit or drains it.

The market has not made this easier. Clients now expect stronger efficiency and clearer value, and they move faster than the old quoting rhythm assumes.

Data point: 66% of services firms say clients are now more demanding but less willing to pay, according to Teamwork.com's 6 Strategic Shifts for 2026 research.

Two shifts matter here. Delivery timelines are compressing, partly because AI now speeds up parts of the work. That leaves a thinner buffer to absorb a soft estimate. Clients are also quicker to question a price and move budget mid-project.

At Teamwork.com, we see that firms rarely miss margin on a single bad quote. More often, margin slips gradually across dozens of projects priced off a gut-feel rate that never counted non-billable time, overhead, or the small favours that erode a retainer.

If any of that sounds familiar, you are not bad at pricing. You are missing the gates that catch these problems before they reach your margin. Treat the main consulting pricing models as a reference, and spend your energy on the method.

The four-gate pricing method, gate by gate

I run every price I touch through the same four checks now, in order. The four-gate method is a sequence of checks, not a single decision: a quote has to clear scope and cost before you even choose fixed-fee or hourly. Each gate answers one question, and a price that skips a gate usually underperforms at month-end.

The order matters: you cannot cost work you have not scoped, and you cannot defend a price you never grounded in real numbers. Skip Gate 1 and every later number rests on sand. Skip Gate 2 and you price on hope. Skip Gate 3 and margin turns into an accident, and skip Gate 4 and the price leaks away in delivery.

The gates also compound. A tight scope makes costing faster. A real cost floor makes the margin decision honest. An honest margin makes the defence conversation simple, because you know exactly what you are protecting and why.

The method works with any tool. Plenty of firms start in a spreadsheet, and that is fine at first. Spreadsheets fail quietly, though, one stale formula or outdated tab at a time. As a firm scales, the gates need to live where the whole team updates them.

Gate 1: lock the scope before you cost a thing

In my experience, the projects that stayed profitable were rarely the ones with the nicest clients. They were the ones whose statement of work said, in plain language, what fell out of scope and what a change would cost.

Scope creep often starts with a small unplanned request: a "quick" extra revision, one more reviewer, a deck "while you're at it". Each favour feels too small to log, so nobody logs it. By month three, the account that looked profitable runs at break-even, with no paper trail explaining why. That is a scoping failure, not a client-behaviour failure.

The fix is a gate. Nobody costs the work until you write down and agree inclusions, exclusions, and a change process.

Good scoping starts with the right questions. I ask what a successful outcome looks like, who signs off, and what happens if a deadline moves. Those answers shape the deliverables, and they surface the risks that belong in the contingency long before they land on the invoice.

You do not need a heavy process. A repeatable intake that captures deliverables, assumptions, and a change path is enough. A shared scoping template beats a blank page every time.

A workable change process has three parts. It states plainly what is included and what you explicitly exclude, sets a rule for what triggers a change request, and agrees in advance how to price that change. For instance, a website build might include five page templates and two revision rounds, and exclude copywriting and post-launch changes. Naming both halves in writing turns a later request into an easy change order rather than a favour.

Get those three on paper and most scope disputes never happen. When scope and change history live in a shared, durable place, a new lead can pick up the account without starting cold. Client-facing teams change, and the projects that survive a handover are the ones with a real paper trail.

For example, say you scope a brand retainer at 40 hours a month for $8,000. Over a quarter, unlogged extras add roughly six hours a month. That is 18 hours delivered and never billed, about $3,600 of value at the retainer's effective $200 an hour. Lock the scope, and those six hours become a change request instead of a write-off.

Gate 2: build the price up from what delivery actually costs

The cost side is where most quotes quietly break, because teams price off a salary number and forget everything stacked on top of it. A $75 hourly salary cost usually lands near $120 an hour once you add benefits, taxes, software, and overhead, and nobody bills 100% of their time, either. Price off the $75 and you start below sustainable delivery cost. Your true cost to deliver is a bottom-up number: loaded labour for every role, plus overhead, tools, and a contingency for what you cannot yet see.

That $27,060 is your floor, not your price. Labour is only 70% of it. The 30% most firms forget is exactly the margin they later lose. Pressure-test your own rate with a billable utilisation calculator.

Utilisation changes your real hourly cost quickly. A designer costing $150,000 a year fully loaded across roughly 2,000 working hours works out at about $88 an hour at 85% billable, and about $115 at 65%. Nothing about the salary changed: time leaked out of the billable column, so the cost per billable hour rose by a third. Two firms with identical rate cards can post very different margins for exactly this reason.

Contingency is not padding; it is priced risk. On a well-scoped build I sit near 10%, and on discovery-heavy work I push it higher. Overhead is the other quiet line. Rent, software, and non-billable roles have to land somewhere, and if they don't sit in your rate, they come out of profit.

There is a client-facing benefit to costing this way. A quote built from real components is easy to explain: hours by role, a clear contingency, and the margin on top. Transparency like that tends to win more work than a round number, because the client can see what they are paying for.

Getting the estimate right is its own discipline, with solid cost-estimation methods worth a separate read. The point for the Cost Gate is simpler. The number must come from real rates and real hours. When SugarCRM unified projects, time, and billing on one platform, they credited less than $20K on over $10M in annual invoicing, per SugarCRM's story. That is what happens when the cost you track is the cost you bill.

Gate 3: set the margin, then let it choose the model

Set the margin before you think about the model. Target margin is the profit you decide to make before you pick how to charge, and it turns your cost floor into a price. If your business needs a 40% gross margin, a $27,060 true cost has to price near $45,000:

$$\text{Price} = \frac{\text{True cost}}{1 - \text{Target margin}} = \frac{27{,}060}{1 - 0.40} \approx 45{,}100$$

The margin sets the price; the model is only how you deliver it.

What margin should you aim for? I'd expect healthy services firms to sit between 40% and 60% gross delivery margin, before overhead outside the project. Below 40%, there is little room to absorb the overruns real delivery produces, so set the number deliberately, as a floor you defend. If hitting it means changing a rate, model the revenue impact across every active project before you commit.

Now model selection finally belongs, as an output. Once you know your cost floor and margin target, the shape of the work points to the model that protects it.

The same $27,060 project behaves differently under two models. As a fixed fee at 40%, you quote around $45,000 and keep the upside if you deliver efficiently, but you own every overrun. On time and materials, you bill loaded cost plus margin per hour, so a growing scope simply bills more.

Our guide to consulting pricing strategies covers the models in depth, so use that reference rather than relitigating them here. The order of operations is what changes with the four-gate method: margin first, model second.

Pricing well is only half the job; you also have to justify the number. When a client pushes back, I anchor to outcomes, not hours, showing what the work is worth to their business and letting the cost floor explain the margin. If a cheaper quote lands on the table, I don't drop the rate; I re-scope, asking which outcome matters most and showing what the lower number leaves out. Framed as an investment in a result, a price lands very differently from a rate defended line by line.

Tiered offers help here too. Giving a client a good, better, and best option tends to move the conversation from "is this too expensive?" to "which version do we want?" Each tier still clears all four gates, so every option protects its own margin.

Gate 4: defend the price while you deliver

I keep seeing the same pattern: the price holds until delivery starts, then it drifts. A project starts on target and still finishes flat, even with a reasonable scope and the right rate, because the margin bled out in twenty small changes nobody tracked. Everyone was heads-down on the work, so nobody saw the fall until the close. That is the Defence Gate failing.

A price is not a one-time decision: 27% of firms say mid-project budget moves are their top frustration, per Teamwork.com's 6 Strategic Shifts for 2026 research. It is a position you hold through delivery, and it needs a mechanism. You track actual cost against estimate as work happens, so an over-budget project shows up in week two.

The mechanism has two moving parts. First, a live budget burn, where your team flags a task at 80% of its budget with half the work left this week. Second, a standing rule that any out-of-scope request becomes a change order before work starts. Good teams treat cost tracking as a live signal, not a post-mortem.

For example, a project budgeted at 300 hours is tracking at 180 hours with only 40% of deliverables done. Caught at the close, that is a loss you explain in a margin review. Caught in week three, it is a change order or a resourcing fix. The difference is simply whether you were watching.

The change-order conversation is easier than most people expect. It works best raised early and anchored to scope, not the relationship. "This wasn't in the brief, here's what it adds and here's the cost" is a normal, professional exchange.

Client reporting should use live delivery data from project reporting, so a monthly update is a byproduct rather than a drain on non-billable time. A standing report gives you the evidence to defend a change order the moment scope moves.

Estimation methods that make the cost gate hold

I lean on bottom-up estimates when scope is clear and three-point estimates when it isn't, because the Cost Gate is only as trustworthy as the estimate behind it. Bottom-up, analogous, parametric, and three-point estimation each have a place. A weighted three-point estimate prices uncertainty quickly:

$$\text{Estimate} = \frac{\text{Best} + 4 \times \text{Likely} + \text{Worst}}{6}$$

Match the method to the moment. Analogous estimates work early, when you only have a rough shape. Bottom-up estimates fit once scope is locked and you can price each task. Three-point estimates earn their keep whenever a piece of work carries real uncertainty.

Rather than re-teach them here, use a dedicated walkthrough and bring the output back to Gate 2. For deeper method guidance, PMI's research on accurate project estimating is a solid external reference.

The pricing mistakes that quietly kill margin

I've made most of these myself, early in my agency career, so none are hypothetical. They are rarely dramatic. They are the same habits, repeated across projects, until they read as normal.

• Pricing from a desired rate down: You pick $45,000, then bend scope to fit, and deliver $60,000 of work for less.

• Treating the estimate as the price: An estimate is your cost floor. Send it out with no margin layer and you price at cost.

• Skipping contingency: Every unknown then comes out of profit. A 5% to 15% buffer is the price of uncertainty.

• Not logging favours: Unrecorded changes cannot be billed or renewed against. If it is not written down, only your margin remembers.

• Never revisiting price: Costs rise yearly. A rate set two years ago is probably underwater now.

• Pricing every project the same way: A tight build and an open discovery carry different risk. Flat-fee both and you lose margin on the messy one.

For example, a firm running 30 projects a year, each leaking a modest $2,000 of unpriced scope, hands back $60,000 a year. That is roughly the loaded cost of a senior hire. The mistakes are small; the total is not, and a gate catches each one before it reaches the invoice.

Each mistake above comes down to a skipped gate. The statements below map them back, so you can see where your own pricing is leaking.

Self-audit: which gate is your pricing missing?

• You quote before scope is written down (Scope Gate).

• You price from a target rate, not loaded cost (Cost Gate).

• You set the margin last, not first (Margin Gate).

• You learn about overruns at month-end (Defence Gate).

Each statement you recognise names the gate to fix. If several apply, start with the earliest, because every later gate rests on it.

How Teamwork.com prices and protects every project

Teamwork.com is the agentic PSA (professional services automation) platform that connects projects, resources, and financials in one place. Generic project tools track the work but cannot manage the money, so margin leaks where the systems meet. Traditional PSAs manage the money, but teams resist using them, so the data going in is unreliable and the numbers are too. Teamwork.com is the best of both: financials and margin visibility captured as delivery happens, in a platform teams actually want to use.

Visibility changes the pricing conversation. When cost and margin are visible as work happens, the gates hold far more easily. The firms that price best are rarely the ones with the cleverest models; they are the ones who can see the truth of a project while it is still moving.

See the true cost before you quote: budgets and time tracking

Guessing at cost is the biggest reason quotes come in underwater, because the number you quote and the cost you incur never meet in time.

I've seen one connected view fix that fast. Budget tracking and time tracking capture loaded cost as work happens, so your next quote reflects what similar work actually took. Over a handful of projects, that history becomes your most reliable estimating input.

Set and watch margin as work happens: profitability reporting

Know your margin while you can still change it. See live project cost and delivery margin in one view, so a project nearing its margin threshold becomes a Tuesday alert and you can act early.

Catch scope creep early: workload and capacity visibility

Scope creep hides in overbooked people before it shows in the numbers. The first signal is workload, not the budget line, and by the time it reaches the budget the margin is gone.

See who is overbooked, underused, or at capacity across every project. A designer at 120% for three straight weeks is a margin problem in plain sight, far cheaper to fix as a resourcing decision than as a write-off. That same visibility makes the client conversation easier, because when an out-of-scope request lands you can show what it displaces and what it costs.

Quote from real actuals: agentic AI

I've found the best guide to the next project is the last one, measured properly. Native AI agents use your actual delivery data to improve utilisation and forecast inputs, so estimates start from evidence, not optimism.

Those agents use the same delivery data your team already works from. So the numbers reflect your real process, not a generic benchmark. A utilisation summary that used to take half a day is ready in minutes.

Across all four gates, the through-line is one shared set of data, captured as delivery happens, in the PSA built for professional services firms.

FAQ