Fixed Price or Time and Materials? Choosing the Right Contract for a UK Software Project
Use fixed price when the scope is genuinely well defined and unlikely to change — a specific integration, a defined feature, a migration with known endpoints. Use capped time and materials when you are building something genuinely new and expect to learn as you go. The deciding question is not which is cheaper, but how confident you are that the spec will survive contact with real users.
Most UK software agencies will quote you either way if you ask. That sounds like flexibility, and it is — but it also quietly moves a significant commercial decision onto you, usually before you have enough information to make it well. Pick fixed price for a project whose scope is still moving and you will spend the engagement arguing about change requests. Pick time and materials for a tightly-defined piece of work and you take on risk you did not need to.
This guide explains what each model actually protects, where each one fails, and the hybrid most well-run UK projects land on.
- The models cost about the same for the same work — what differs is who carries the risk of being wrong about scope.
- Fixed price carries a 15–30% contingency premium. You are buying budget certainty, and paying for it.
- Fixed price on an unclear scope does not remove uncertainty — it converts it into a change-request process that makes flexibility expensive.
- The structure that works most often is fixed-price discovery first, then a build priced on what discovery found.
- Ask what contingency is in the number, what is excluded, and who owns the code. Vague answers to any of the three are the signal.
What each model really does
Both models cost roughly the same for the same work. What differs is who carries the risk of being wrong about scope.
- Fixed price moves scope risk to the agency. They quote a number, and if the work takes longer than estimated, that is their problem. In exchange they price in a contingency buffer — typically 15–30% — and they will defend the original scope, because every change eats their margin.
- Time and materials keeps scope risk with you. You pay for the time actually spent, so there is no buffer built in and no incentive to resist change. In exchange you carry the uncertainty: if the work takes 30% longer, you pay 30% more.
- Capped time and materials splits it. You pay for actual time, but with an agreed ceiling the agency cannot exceed without your approval. You get the cost transparency of T&M with a hard limit on exposure.
| Model | Who carries scope risk | Best for | How it fails |
|---|---|---|---|
| Fixed price | The agency | Defined integrations, completed designs, a known slice of an existing system | Change becomes a negotiation; the agency is paid to defend the original scope |
| Time and materials | You | Undocumented legacy work, genuine R&D, scope that depends on an undecided external factor | Open-ended spend with no ceiling if reporting and prioritisation are weak |
| Capped time and materials | Shared | First versions and new products where the shape is still being learned | The cap becomes a de facto fixed price if the backlog is not genuinely prioritised |
When fixed price is the right call
Fixed price works when the unknowns have already been resolved. In practice, that means:
- The integration is specified. You are connecting to an API with published documentation, known endpoints and a sandbox you can test against.
- The design is done. Someone has already produced the screens, and they have been reviewed by the people who will actually use them.
- The work is a defined slice of a system that already exists, rather than a new product whose shape is still being argued about internally.
- You need budget certainty for a board or a grant — sometimes the commercial requirement genuinely outranks the technical one, and that is a legitimate reason to accept the contingency premium.
The pattern across all four: the spec is unlikely to change, because the learning has already happened. If that is not true of your project, fixed price does not remove uncertainty — it just converts it into a change-request process.
When time and materials is the right call
T&M is the honest model for genuinely new work, and the usual signals are easy to spot:
- You are building a first version and expect real users to change your mind about at least one significant feature.
- The legacy system is undocumented. Nobody can tell you exactly what it does, so nobody can honestly quote a fixed price to replace it.
- Requirements depend on something not yet decided — a partner contract, a regulatory position, a pricing model still in discussion.
- You want to be able to change direction after a demo without renegotiating the contract each time.
A fixed price on an unclear scope is not a guarantee. It is a bet that the discovery you have not done yet will agree with the estimate somebody wrote before doing it.
The hidden cost of fixed price on unclear scope
The failure mode is predictable and expensive. You agree a fixed price against a spec written before anyone truly understood the problem. Four weeks in, a real user sees the first working build and says something reasonable that was not in the spec. Now every improvement is a commercial negotiation rather than a technical decision, and the agency is incentivised to say no — not out of bad faith, but because they priced the original scope and absorbing changes costs them money.
The result is a product that matches a document rather than a product that works. Teams often describe this as the agency being inflexible, when the real cause was choosing a contract model that made flexibility expensive for everybody.
What most well-run projects actually do
The structure that works most often is a two-stage engagement, and it is worth asking for by name:
- Stage one: a fixed-price discovery. Two to four weeks, a defined deliverable — technical architecture, screen designs, a prioritised backlog, and a realistic estimate. Small enough to fix a price on honestly, because the scope is “produce these documents”, not “build this system”.
- Stage two: priced from what discovery found. With the unknowns resolved, the build can be fixed price if the scope is now genuinely stable, or capped T&M if it is not. Either way the number is grounded in analysis rather than optimism.
This also gives you a low-cost exit. If discovery reveals the project is bigger than the budget, you have spent a few thousand pounds finding out rather than committing six figures to something that was mispriced from the start.
Questions worth asking before you sign
- What happens when we want to change something? A good answer describes a specific process, not “we will be flexible”.
- What contingency is in this fixed price? Agencies that will not answer are usually carrying a large one.
- On T&M, what do I see weekly? You should get hours by person and task, not a single monthly invoice figure.
- What is explicitly excluded? Data migration, training, third-party licences and post-launch support are the four most commonly left out of a headline number.
- Who owns the code? It should be you, on payment, in writing.
Frequently asked questions
Is fixed price more expensive than time and materials?
Usually yes, for the same delivered scope — typically 15–30% more, because the agency prices in contingency for the risk it is absorbing. You are paying for budget certainty. Whether that premium is worth it depends entirely on how likely your scope is to change.
Can we start fixed price and switch to time and materials?
Yes, and the cleanest version is planned from the outset: a fixed-price discovery phase followed by a build priced on what discovery found. Switching mid-build is possible but usually signals the original scope was wrong, so expect a renegotiation rather than a simple change of model.
What is a capped time and materials contract?
You pay for time actually worked, but with an agreed ceiling the agency cannot exceed without your written approval. You keep the transparency and flexibility of time and materials while capping your worst case, which is why it suits products whose scope is still genuinely being learned.
How do I stop a time and materials project from overrunning?
Insist on a cap, weekly reporting of hours by person and task, and a prioritised backlog so the most valuable work is always done first. If budget runs out, you should have a working product missing its lowest-priority features — not a half-finished one missing critical ones.
Which model does Autratech use?
Both, and we will recommend one rather than offering both and letting you guess. A tightly-specified integration or feature suits a fixed price; a genuinely new product suits a capped time-and-materials retainer. Every engagement starts with a scoping call and a short technical discovery, and you approve scope before we write a line of code.
The practical takeaway
Do not choose the contract model first. Work out how stable your scope genuinely is, and let that decide. If you cannot answer that yet, a short paid discovery is almost always cheaper than guessing — in either direction.
If you want a straight recommendation on which model fits something you are planning, tell us what you are building. You can also see how we structure engagements on our pricing page.
