Fixed price vs time and materials vs value pricing: an honest comparison

The three pricing models aren't better and worse versions of each other. They distribute risk differently, and that's the entire decision. Fixed price puts delivery risk on you. Time and materials puts it on the client. Value pricing tries to move the conversation off cost altogether, and only works when you can name what the outcome is worth.

Everything else people say about them is downstream of that.

The same project, three ways

A hypothetical project estimated at 300 hours at a loaded cost of $95 an hour — so $28,500 of delivery cost if the estimate is right.

Fixed price at $57,000. That's an implied rate of $190 an hour and a 50% margin if it lands on estimate.

Time and materials at $150 an hour. Margin per hour is ($150 − $95) / $150 = 36.7%, and it stays at 36.7% whatever happens.

Value price at $85,000, based on what the client says the outcome is worth to them rather than on hours.

Now run all three against three delivery outcomes: on estimate (300 hours), 27% over (380 hours), and 52% over (456 hours).

Hours delivered	Fixed $57,000	T&M at $150/h	Value $85,000
300 (on estimate)	$28,500 cost → 50.0%	$45,000 revenue → 36.7%	$28,500 cost → 66.5%
380 (27% over)	$36,100 cost → 36.7%	$57,000 revenue → 36.7%	$36,100 cost → 57.5%
456 (52% over)	$43,320 cost → 24.0%	$68,400 revenue → 36.7%	$43,320 cost → 49.0%

Three things stand out.

The T&M column doesn't move. That's the whole feature: your margin is set by the gap between your rate and your cost, and overruns are the client's problem.

Fixed price is better than T&M when you're right and worse when you're wrong. At 380 hours the two models pay exactly the same. Past that, fixed price is losing you money that T&M would have earned.

Value pricing is the most forgiving. Even at 52% over — a genuinely bad overrun — it's still ahead of a perfectly-executed T&M engagement. Price decoupled from hours absorbs execution error.

What each one is actually for

	Fixed price	Time and materials	Value
Risk sits with	You	Client	You
Needs	A tight, stable scope	Trust and visibility	A quantifiable outcome
Client sees	Certainty	Fairness	A business case
Fails when	Scope moves	Client won't accept open-ended cost	You can't name the value
Rewards	Efficiency	Nothing in particular	Expertise and speed

The reward row is worth sitting with. On T&M, getting faster reduces your revenue. If your team's skill means you deliver in 180 hours what a competitor takes 300 hours to do, hourly billing punishes you for it — you bill $27,000 where they bill $45,000, and the client thinks they're the better value.

That's not an argument against T&M. It's an argument for pricing outputs rather than inputs anywhere your work is genuinely faster than the market.

Fixed price without the downside

Most fixed-price disasters are scope failures, not estimation failures. Three adjustments remove most of the risk:

Fix the scope, not the project. Price phase one fixed, then price phase two when phase one has told you what phase two is. A discovery phase priced at $8,000 that de-risks a $60,000 build is the cheapest insurance in agency work.

Put the buffer in the hours, not the margin. If you add 20% contingency and then discount 20% to win the deal, you've sold the buffer.

Write down the assumptions the price depends on. Fixed price means fixed for the scope described, and the SOW should say which assumptions, if false, trigger a change request.

When clients say "we don't do time and materials"

Usually this means "I've been burned by an open-ended invoice", not "I insist you carry all the risk". Two structures usually get you there:

Capped T&M. Billed hourly, with a ceiling. The client can't be surprised; you still get paid for what you do up to the cap. You are giving up the upside of finishing early, which is a real concession, so don't set the cap at your estimate — set it at your estimate plus the buffer.

Retainer with a scope band. A fixed monthly fee for an agreed volume of output, with a stated rate above the band. This is how most content, SEO and social engagements should be priced, and almost none of them are.

Making value pricing real

Value pricing collapses the moment you can't answer "worth how much?" with a number the client recognises. Some engagements have one available — a campaign tied to a revenue target, a conversion improvement, a process that removes a cost. Many don't, and pretending otherwise just means charging a big number and hoping.

The usable test: can the client, in their own words, complete the sentence "if this works, it's worth roughly ___ to us"? If they can, you have a basis. If they can't, price the work and stop performing.

Where value pricing does apply, don't anchor on hours anywhere in the proposal. A value price next to an hours breakdown is just a fixed price with an embarrassing markup.

A default worth adopting

Most agencies should use fixed price for work they've done before with a scope they control, capped T&M for anything exploratory, and value pricing only where the client has already named a number.

Pick your next three proposals and assign a model deliberately rather than by habit. Then run the table above on each — on estimate, 27% over, 52% over — and look at the worst cell. If you can't live with it, you've picked the wrong model, and it took ten minutes to find out.