Time and materials vs fixed fee: how to choose

Choose fixed fee when the scope is genuinely fixed. Choose time and materials when it is not. Everything after that is detail.

A fixed fee contract (also called fixed price) settles scope, price and date before anyone writes code, so moving any one of the three means renegotiating the contract. A time and materials contract bills the hours the work takes, so scope can move without a new contract and you learn the total at the end rather than the beginning.

The test is not project size, and it is not budget. It is whether you can write down what done looks like in enough detail that a stranger could build it and you would accept the result. If you can, a fixed fee moves the estimation risk onto the vendor, and that transfer is worth paying for. If you cannot, a fixed fee does not remove the risk. It converts it into change requests, and you pay for it in argument instead of in hours.

There is a third answer the two-way comparison hides. Some work has no finish line, and neither contract is the right shape for it. A team billed monthly per member is, and it is the third model we run.

What is a time and materials contract?

A time and materials contract bills the hours the work takes, at an agreed hourly rate, instead of pricing a scope in advance. Scope can change without a new agreement. The budget is known at the end, not the beginning. Dates are forecasts, not commitments, because the thing forecast moves. You carry the estimation risk and get the steering wheel.

Its obvious failure mode is that a vendor billing by the hour has no commercial reason to be quick. So the thing worth testing is not the rate. It is what a vendor does when the honest answer is worth less to them than the slow one. Boruch Akbosh, founder of Mover Technologies, describes how our relationship started in his Clutch review:

“I was getting 30–40-hour quotes, but then Artelogic offered to fix my problem in 30 minutes without upgrading to .NET. I accepted the offer and continued to work with Artelogic from there. I liked the fact that they were honest, offered good quotes, and had skill sets right down my alley.”

The same Clutch review set out the billing rule the engagement ran on:

“I’m not sure how Artelogic works with other clients, but if there is a fix, it’s never billable. We are only billed for additional functionalities, logic changes, and so on.”

That is the term worth negotiating for in any time and materials contract: defects are the vendor’s cost, new work is yours. Without it, the model quietly pays for rework. The build is written up as a freight management platform for a logistics startup, one of the freight and moving platforms behind our logistics and transportation software work, and it began under our former Artelogic brand.

What is a fixed fee contract?

A fixed fee contract prices an agreed scope at an agreed number for an agreed date. All three are settled before the work starts, and none of them moves without a renegotiation. It transfers the estimation risk to the vendor, which is the whole reason to want one, and it buys that transfer with a longer planning phase up front.

What that phase feels like from the client’s side is worth knowing before you agree to it. Patrick McGuire, IT director at Barrett Values Centre, described the start of a report-automation build in his Clutch review:

“We started with a kickoff to get the scope of the work. We then shared our mockups and sample data. After a week or 2, they came back with a scoped estimate with hours for each task.”

A week or two of specification before a line of code is what the certainty costs. It is also what makes the certainty real, because an estimate built that way is one you can hand to other people. The same Clutch review says why:

“The work progression matched the estimate that they provided. This was nice because we were able to confidently share expected deliverables with other stakeholders.”

That build is written up as report automation for a culture analytics company.

How the three models compare

The three differ on five things a buyer actually weighs: how freely scope can move, how the budget behaves, how the deadline behaves, what the model asks of you, and where each one fits. In practice one row decides the contract and the rest follow, so read the one that matters most to you first.

What you are weighing	Time and materials	Fixed fee	Team extension
Scope flexibility	Changes any week, with no new contract	Frozen at signature; every change renegotiates price and date	Whatever your backlog says this month
Budget behaviour	Known at the end; you control it by controlling scope	Known at the start, and it holds as long as the scope does	A predictable monthly amount that moves only when the team size moves
Deadline behaviour	A forecast that moves with scope	A commitment the vendor carries	Set per release by whoever owns the plan
What it asks of you	Decisions on demand, because the model bills the waiting as well as the building	A specification a stranger could build from	Direction and priorities, continuously
When it fits	Requirements still moving: modernization, discovery-heavy builds, anything shaped by user feedback	A specified deliverable with a hard date: a compliance change, an integration, a designed release	Work with no finish line: a product you will keep running and extending

When fixed fee is the right call

Fixed fee earns its place when you can specify the result and you need the date more than the flexibility. Four situations qualify: a deliverable with a regulatory or contractual deadline, a migration or integration with a known target, a release you have already designed, and any piece of work small enough that a week of specification is proportionate.

It stops earning its place the moment your own organisation cannot hold still. If three stakeholders will see the design for the first time after signature, you have not bought certainty. You have bought a change-request process with a price list attached. The same applies when the requirement depends on something you do not control yet: a partner API, a regulator’s final wording, a board decision. Fix the price of what is known and leave the rest outside the contract.

Where the scope really is fixed, the discipline shows up as a number that does not move. Kseniya Savelyeva, marketing director at Perseus Group, in her Clutch review:

“All we can ask for from a development team is to do exactly what we want, on time and on budget. This is what Artelogic did, for every little feature we wanted. They did it a week early, giving us extra time to flesh out minor details. The cost remained the same, so there was no extra charge for my pickiness.”

Read the last sentence as a scoping result rather than a favour. The pickiness fitted inside the scope because the scope had been specified well enough to contain it. That work was done under our former Artelogic brand.

When time and materials is the right call

Time and materials earns its place when the requirement is still being discovered, when the system already exists and you are changing it a piece at a time, or when the roadmap depends on what users do next. In all three, a specification written today is fiction, and a fixed price built on fiction is one you renegotiate.

Legacy modernization and integration work are the clearest instances: you cannot specify the fix for behaviour nobody has finished mapping, and the real scope shows up as you open the system.

Buyers state this about themselves more plainly than vendors do. An anonymous marketing director at a Toronto digital marketing firm gave the selection criterion in their Clutch review:

“Some of our requirements were still evolving, so the selected firm had to be flexible.”

Flexible is not the same as open-ended, and the same Clutch review names what keeps the difference visible:

“They were very transparent about their deliverables, and their estimates were very thorough.”

“They put an additional developer on the project when needed. They stuck to the budget and timeline. When I was concerned about missing the deadline, they reassured me.”

A time and materials engagement that cannot tell you where the hours went is not flexible. It is unaccountable. Ask for the estimate per task, the burn against it, and the right to stop at the end of any sprint. With those three the model behaves; without them the flexibility is all on the vendor’s side.

Vendors commonly offer a variant of it, and it is worth recognising when you are handed one. Capped time and materials, usually written as not-to-exceed, bills by the hour like ordinary time and materials but sets a ceiling the total cannot pass without a new agreement. It behaves like time and materials until the cap comes into view and like a fixed fee after that, so the flexibility quietly ends when the budget does. Read the cap as the real scope boundary rather than as a discount, and pair it with the same three asks above: without the per-task estimate, the burn against it and a clean stopping point, a cap only tells you when the argument starts.

When you should buy a team instead

Some work never finishes. A product you will keep running, a platform that keeps absorbing features, a backlog that regenerates faster than you close it. Pricing that as a series of projects means writing a new contract every quarter and pretending each one is the last.

The third model is team extension: engineers who join your delivery process, billed monthly per team member rather than quoted against a scope. The budget becomes a line item you can plan around. Scope stops being a contractual object at all, because what you are buying is capacity rather than a deliverable. On a dedicated team or team extension there is no recruitment or placement fee, and nothing separate is charged for sourcing, vetting or placing an engineer.

Andrew Taylor, head of IT at JewelCandle, describes the elasticity in his Clutch review:

“Artelogic scales up or down based on our needs. For example, if we need to create a big feature, they quickly onboard resources. Their team performs all of the pre-interviews and send us qualified candidates. When we finish a large project, their team scales down.”

“We typically work on projects for 3–6 months and then scale down. Once there’s another project, we scale up our resources again.”

Eight of our 61 public Clutch reviews describe scope that grows after the first project: engagements that expanded, or teams the client kept on longer than planned. That pattern is what this model exists for. The JewelCandle work, done under our former Artelogic brand, is written up as ecommerce software development for a European manufacturer.

How you pay is a different question from who owns delivery. If you are also deciding whether the vendor runs the plan or you do, managed delivery vs team extension works that one through. If you have not yet settled whether to build the capability internally at all, start with in-house vs outsourcing and come back here afterwards.

What no contract model fixes

No billing model makes a date immune to a change of mind, and the honest version of a slipped deadline usually starts on the client’s side. An anonymous director at a Singapore design studio put the mechanism plainly on Clutch:

“Deadlines sometimes need to move out, though not necessarily at their fault. It can be that a client changes the scope. I really can’t fault them on it. Deadlines are always met though.”

Our own record is not spotless either, and a client who runs nine vendor teams counted it. An anonymous CEO of a software consulting firm in Nashville, on Clutch:

“We have had six out of seven on time and on budget project executions. Each engagement was a minimum of six months of effort. They were successful six out of seven times. That’s pretty good.”

Six of seven, not seven of seven. The same client, in the same Clutch review, on how that sits against the rest of the bench:

“I work with nine different teams. Their team is one of three that I never have to worry about the quality of their deliverables. They’re probably my second most reliable team when it comes to quality.”

What makes either model checkable is a reporting cadence you can hold the estimate against. Igal Nikomarov, owner of Matok Realty, in his Clutch review:

“As far as meeting deadlines, they provided a workflow report and targets were met on or before their estimated deadlines, which was good.”

And Jeff Sides, VP of IT at MiniMoves, in his Clutch review, on the .NET order management system we took over for that moving company:

“They’re good at staying within cost estimates.”

That work is written up as the MiniMoves order management platform. Ask for the same reporting in writing whichever model you sign. A contract model allocates risk; it does not remove it.

How unicrew prices engagements

We run three engagement models. Time and materials is billed hourly and quoted per project, fixed price is outcome based and quoted per project, and team extension is billed monthly per team member. Which one we propose depends on the work in front of us, and we will say which on the call.

We publish the model rather than a number, because a figure quoted before anyone has seen the shape of the work is a figure that is wrong for most engagements. On a dedicated team or team extension there is no recruitment or placement fee, and nothing separate is charged for sourcing, vetting or placing an engineer. Most engagements start within two to four weeks.

What sits behind that: 120+ projects across 12 countries since 2012, and 100+ senior in-house engineers across six countries. Of our 61 public Clutch reviews, 34 name on-time or on-budget delivery: deadlines the client says we met, or budgets they did not have to renegotiate. The count is computed on our client reviews page, where you can filter by that claim and read every review behind it. Two of the 34:

“unicrew meets their deadlines and stays on budget. They’ve finished the API on time, even with tight deadlines.”

An anonymous CTO at a science tech company in Germany, on Clutch. That work is written up as a carbon footprint SaaS platform for a climate tech startup.

“The Artelogic team finished the project on time and within budget. Their work was of excellent quality, making the cost justified.”

Tal, CEO of a cloud solutions company in Israel, on Clutch, about work done under our former Artelogic brand.

If you have already chosen a model, the service page is the next step: dedicated software development team and managed teams both set out how a team is assembled and what you sign. If you have not chosen, tell us the shape of the work and we will say which model we would propose, and why.