Fixed fee or time and materials: what each one makes your vendor optimize for
Neither model is honest or dishonest. Each one rewards a different behavior, and knowing which behavior you are paying for is the whole decision. Plus the two hybrids that actually work.
By Laurent TulpanPublished 9 min read
The argument is usually framed as trust. It is not about trust. Both models are honest, and each one pays a vendor to behave in a specific way. Knowing which behavior you are buying is the entire decision.
Time and materials pays for attention
You buy hours or days. Whatever happens, they are billed.
What it rewards. Thoroughness, and being needed. Nobody is punished for a task taking longer than expected, which is exactly right when nobody could have known how long it would take. It also means there is no financial mechanism pushing the work toward a finish.
Where it is the correct model. Genuinely undefinable work. Investigating an intermittent failure whose cause is unknown. Reverse-engineering a system nobody documented. Supporting a team through a migration whose surprises are the point. In all three, a fixed price would be a bet rather than a commitment, and you would pay for the bet.
What it hides. The total. An estimate under time and materials is not a commitment, and everybody involved knows it, which is why the conversation about scope happens in month four rather than month zero.
The two clauses that make it safe. A ceiling, in days or in money, beyond which work stops pending your written approval, obtained before the work rather than after. And a weekly figure of days consumed against days remaining, sent without being asked. Neither slows the work down. Both mean the overrun conversation happens while there is still room to act.
Fixed fee pays for finishing
You buy an outcome: this scope, working, by this date, for this price.
What it rewards. Closure. The vendor absorbs the overrun, so the vendor wants a tight scope, explicit acceptance criteria and no surprises. Those are the same three things you want, which is unusual enough to be worth noticing.
What it costs. A visible risk premium. A vendor carrying the overrun prices the possibility of it, and that is not padding: it is the price of transferring the risk, exactly as an insurance premium is not a rip-off because the fire did not happen.
Where it fails. On a scope nobody can define. A fixed price on unclear scope produces one of two outcomes: a vendor who did not understand and will fight every request as out of scope, or a vendor who understood and priced a cushion so large you pay for the uncertainty twice.
What it requires from you. Acceptance criteria you can state before work starts, and discipline about changes. The model’s weak point is not the price, it is the change request: every fixed-fee engagement lives or dies on how the first scope change is handled. Agreeing the mechanism in advance, in writing, is worth more than negotiating the price down.
The two hybrids that actually work
Scoping paid separately, then a fixed fee. A short paid engagement produces the specification, the estimate and the acceptance criteria. That deliverable belongs to you whether or not you continue with the same vendor, and that condition is what makes it honest: a scoping study you cannot take elsewhere is a sales document with an invoice attached. Once the scope exists, the build is fixed. This is how we work, and the technical due diligence page is that first step.
Fixed fee on deliverables, bounded commitment on the leadership. Because scoping, arbitration and deciding what not to build have no fixed shape. Both halves carry a stated end, so nothing continues by default. That distinction is the same one that separates the three ways a fractional CTO gets priced.
Choosing, in three questions
Can I write down what counts as done? If yes, fixed fee, and you are leaving money on the table if you accept time and materials. If no, the honest answer is not to pick a model, it is to pay for a short piece of work that makes the answer yes.
Who is better placed to absorb an overrun? A vendor absorbs it across a portfolio of engagements. You absorb it out of one budget. That asymmetry is the argument for fixed fee, and it is also why a vendor who refuses to commit on a definable scope is telling you something about their confidence in their own estimate.
What happens to the relationship if the estimate is wrong? Under time and materials, you pay and resent it. Under fixed fee, they absorb it and may cut corners at the edges. Both failure modes are real. The clause that addresses both is acceptance criteria specific enough that corner-cutting is visible, which is a property of the requirements document rather than of the pricing model.
Why we do not sell open-ended day rates
Not on principle, and not because time and materials is dishonest. Because we have watched what it becomes on both sides.
The client stops asking whether they still need the arrangement, since it is a line item rather than a decision. The vendor stops asking whether they are still adding anything, since the invoice goes out either way. Nobody is acting badly, and eighteen months later there is a dependency neither party chose.
A fixed fee forces the question that matters at the outset: what exactly are we buying, and how will we know we got it. That conversation is uncomfortable, it takes an extra week, and it is the single highest-return week in the whole engagement.
Where a client’s procurement rules impose time and materials, we adapt, with a written ceiling and a stated exit objective. Refusing on doctrine would be pointless. What we do not do is let an arrangement run without either.
Further reading
- What a fractional CTO costs: the three pricing models, for the leadership half
- Writing requirements that produce comparable quotes: what makes a fixed fee possible
- Acceptance criteria that end the argument: the clause that protects both sides
- Build or buy: the decision upstream of the contract
- Technical due diligence: the paid scoping that makes a fixed fee honest
The questions this raises, answered straight.
Which model is cheaper?
Time and materials is cheaper on the quote and frequently more expensive at the end, because the quote is not a commitment. Fixed fee includes a risk premium you can see, which is the price of the vendor carrying the overrun instead of you.
The comparison that matters is not the two numbers, it is who pays when the estimate turns out to be wrong.
Why do vendors prefer time and materials?
Because it removes their exposure, and for scopes that genuinely cannot be defined, that is a legitimate position rather than an evasive one.
It becomes a warning sign when the scope is definable and the vendor still refuses to commit: what they are telling you is that they are not confident in their own estimate, which is information you should act on.
Can a fixed fee cover an unclear scope?
No, and a vendor who says yes has either not understood the scope or has priced a cushion large enough that you are paying for the uncertainty twice.
The correct answer to an unclear scope is a short paid piece of work that clarifies it, priced on its own, after which the rest can be fixed. Two contracts, not one heroic guess.
What about success fees or shared risk?
They work when the outcome depends almost entirely on the vendor and can be measured without argument, which is rarer than it sounds in software.
Most systems succeed or fail partly on adoption inside your organization, which the vendor does not control. Where the measure is contestable, a success fee mostly buys you a dispute at the end instead of a disagreement at the start.
Recognize your situation? 20 minutes is enough to tell.
Describe what is happening on your side, and where your data and customers are. You get a plain answer on whether we can help, and a better route if there is one.