Fixed price feels safer because the number is known. That feeling is the point of it, and it is frequently wrong. Time and materials feels riskier because the number is not known, which is honest but hard to take to a board.
Neither model is better. They allocate risk differently, and the right choice depends on one question: how settled is the work, really?
What Each Model Actually Means
Under fixed price, the supplier agrees a defined deliverable for an agreed sum. They carry the risk that it takes longer than expected, and they price that risk into the number. You carry the risk that the thing you specified is not the thing you needed.
Under time and materials, you pay for the effort spent at an agreed rate. You carry the delivery risk, and in exchange you can change direction without renegotiating.
Everything else, capped models, milestone billing, retainers, is a way of splitting that same risk somewhere between the two.
When Fixed Price Is Right
Fixed price works when the requirements are genuinely settled and both sides understand them the same way. A marketing site with signed-off designs. A well-defined integration between two systems whose interfaces are documented. A migration where the source and target are both known.
The common thread is that discovery has already happened. Someone has done the thinking, and what remains is execution.
It also works well for a discovery phase itself. Fixing the price of two weeks of investigation is low risk for everyone and produces the information needed to price the rest properly.
How Fixed Price Fails
The failure is rarely that the supplier loses money. It is that both sides start defending their position instead of building the product.
Once the price is fixed, every change becomes a negotiation. The supplier is financially punished for flexibility, so they stop offering it. You learn something in week six that should change the design, and the honest response, that this is new information and the plan should change, now costs money and goodwill. So it does not get raised.
The second failure is quality you cannot see. When margin is under pressure the things that get quietly dropped are the invisible ones: tests, error handling, documentation, accessibility. The demo still works. The cost arrives eighteen months later in maintenance.
The third is padding. A supplier pricing an uncertain scope must add contingency or go out of business. On work that then goes smoothly, you have paid for risk that never materialised.
When Time and Materials Is Right
Time and materials suits work where you expect to learn something that changes the plan: new products, anything user-facing where the first version is a hypothesis, research-heavy work, and ongoing development of a system that is already live.
It also suits work where speed matters more than certainty. Nothing slows a project down like a change control process.
The trade is that it demands more of you. Someone on your side has to prioritise, make decisions quickly and watch the burn. Time and materials with an absent client is the most expensive model there is.
Controlling the Risk in Time and Materials
Agree a cap, and treat reaching it as a decision point rather than a failure. A not-to-exceed figure with a commitment to flag at seventy per cent gives you most of the protection of fixed price without the rigidity.
Bill and review in short cycles. Fortnightly is usually right. A month is long enough for a misunderstanding to become expensive.
Insist on working software at the end of each cycle, not a progress report. A demonstration is difficult to fake; a percentage complete is not.
Watch the rate of useful change rather than hours logged. Hours tell you what was spent. Only the demo tells you what was bought.
The Models In Between
Discovery then fix. Buy a short, fixed-price discovery, then price the build properly with the information it produces. This is the sensible default for most substantial projects, and it converts the biggest unknown into a small, bounded spend.
Capped time and materials. Pay for effort, with a ceiling. The supplier carries the tail risk, you keep the flexibility.
Fixed capacity. Buy a team for a period rather than a scope. Suits ongoing product work where the roadmap is expected to move.
Fixed price per increment. Each phase is fixed once its scope is understood, rather than fixing the whole programme up front. Keeps the commercial clarity while letting the plan respond to what the last phase taught you.
Not sure which model fits your project?
Tell us what you are trying to do and we will tell you honestly whether we are the right people for it.
Get in touch