Budgeting

Fixed price vs time and materials: which contract to use

Both pricing models fail in predictable ways. How to tell which one fits the work you actually have, and how to structure either so incentives point in the same direction.

3 min readBy Mayur Kale

Key takeaways

  • Fixed price transfers risk to the supplier, and they price that risk into the number. You always pay for it, the question is whether you get the certainty you paid for.
  • Time and materials is cheaper when scope is genuinely unknown and more expensive when it is used to avoid making decisions.
  • The best structure for most work is a sequence of short fixed-price milestones, re-scoped between each.
  • Watch the incentives: fixed price rewards doing less, time and materials rewards taking longer. Both are managed by short cycles and visible output.
  • Any contract without a written non-goals list will produce a scope argument. This is the single highest-value clause you can add.

Buyers want fixed price because it feels safe. Suppliers want time and materials because it feels fair. Both positions are rational, and both models fail in ways that are entirely predictable, which means you can plan around them.

What each model actually does

Fixed price transfers risk to the supplier. They agree to deliver a defined thing for a defined number, and they absorb the overrun if it takes longer.

That risk is not free. Any competent supplier prices in a buffer, so you are paying a premium for certainty. This is a legitimate trade, insurance is a reasonable thing to buy, but understand that you are buying it.

Time and materials keeps risk with you. You pay for effort. If it takes longer, you pay more. The supplier's margin does not depend on estimating well, so there is no buffer in the rate.

Neither is dishonest. They allocate risk differently, and the right choice depends on who is better placed to bear it.

The failure modes

Fixed price fails when scope was unknown at signing. The price was set when least was known, so both parties spend the project arguing about whether things are in scope. The relationship becomes adversarial at exactly the point it needs to be collaborative, and the supplier's rational move is to deliver the minimum defensible interpretation.

Time and materials fails when nobody is deciding. Without a forcing function, scope expands gently, everyone stays busy, and six months later there is a lot of work and no product. This is rarely malice, it is the absence of a decision point.

What works for most engagements

A sequence of short fixed-price milestones, re-scoped between each.

  • Each milestone is three to six weeks
  • Each has a written specification with non-goals and acceptance criteria
  • Each ends with something deployed
  • Price is fixed for the milestone, and only for the milestone
  • Between milestones, either side can change direction or stop

This gets you the certainty of fixed price on a horizon short enough for the estimate to be real, and the flexibility of time and materials at the boundaries. It also creates a natural exit at every milestone, which is worth a great deal when a relationship is not working.

Fixing scope for four weeks is realistic. Fixing it for six months is fiction, and everyone signing knows it.

When to use each in its pure form

SituationModel
Well-defined project, stable requirementsFixed price
Discovery or scoping workFixed price, small
Exploratory product developmentTime and materials, capped
Ongoing team augmentationTime and materials
Migration with a clear end stateFixed price
Anything where you cannot describe "done"Time and materials, or do discovery first

That last row is the important one. If you cannot describe what done looks like, a fixed price is not protection, it is a guess with a signature on it, and the supplier will resolve the ambiguity in their favour because they have to.

Structuring either well

Whichever you choose:

Write the non-goals down. The most valuable clause in the document. Most scope disputes are about unstated assumptions, not broken promises.

Require deployed increments. Not status reports. Working software every week or two is the only reliable signal of progress, under either model.

Define acceptance concretely. Specific enough that two people would independently agree on whether it had been met.

Agree how change is handled. A written change with a price attached, decided explicitly. Not "we'll be flexible" and not "the contract is the contract".

Cap time and materials by phase. A ceiling with a review at the boundary keeps flexibility without giving up control.

What to actually optimise for

The pricing model is a smaller lever than founders assume. What determines whether an engagement goes well:

  1. Whether scope was genuinely resolved before building started
  2. Whether working software appears frequently enough to catch a wrong direction
  3. Whether both sides can walk away at a defined point without disaster

Get those right and either model works. Get them wrong and neither does.

Frequently asked questions

Is fixed price or time and materials better for software development?

Fixed price suits well-understood work with stable requirements and a clear definition of done. Time and materials suits exploratory work where the requirement is discovered by building. Most real projects contain both, which is why a sequence of short fixed-price milestones, re-scoped between each, outperforms either pure model for the majority of engagements.

Why do fixed-price software projects go wrong?

Because the price is set when the least is known. The supplier prices in a risk buffer, and from that moment both sides are arguing about scope rather than solving the problem, every change becomes a negotiation. It works well when the scope is genuinely nailed down beforehand, which requires paid discovery, and poorly when it is used to avoid doing that discovery.

How do I stop a time-and-materials project from running away?

Cap it and review often. Set a budget ceiling per phase, require a deployed increment every week or two, and hold a go/no-go decision at each phase boundary. The risk in time and materials is not dishonesty; it is drift, and drift is only visible if you look at working software rather than status reports.

What is a discovery or scoping engagement?

A short paid piece of work, usually one to two weeks, that produces a specification, an architecture outline and a milestone plan. It is what makes an honest fixed price possible on the work that follows, and it should be structured so the output is useful even if you take it to a different supplier.

Should the specification be part of the contract?

Yes, along with an explicit non-goals list. Most scope disputes are not about what was promised but about what both sides assumed without saying. Writing down what is excluded converts those assumptions into decisions before anyone has spent money on them.

References & further reading

  1. [1]
  2. [2]

Turn the trend into shipped software

I contract as a senior fullstack and data engineer, building the production systems your customers and your reporting depend on.

Explore services↗