Fixed price vs time & materials: which contract protects you?
Time and materials shifts delivery risk onto you; a fixed price shifts it onto the vendor. Oxford researchers who studied 1,471 IT projects found an average cost overrun of 27% — and one project in six became a “black swan,” running 200% over budget. Which model fits your project comes down to one question: can the scope be fixed honestly before the build starts? Here is how to decide.
What each model actually means
Under time and materials (T&M), you pay day rates until the work is done — however long that takes. Under fixed price, you sign a scope, a price, and a date, and the vendor delivers against all three. Neither model is inherently better; they allocate the same risk differently. On T&M, every unclear requirement, every underestimate, every slow week is your cost. On fixed price, it is the vendor's. The rest of the decision follows from who is better placed to carry that risk on your specific project.
The incentive problem with day rates
Hourly billing makes ambiguity profitable: an unclear requirement is not a problem to remove, it is more billable hours. That is one reason overruns are normal rather than exceptional — across 1,471 IT projects, Flyvbjerg and Budzier measured a 27% average cost overrun, with one in six projects running 200% over budget and almost 70% over schedule. The Project Management Institute's survey of 5,402 professionals adds that 52% of projects experience scope creep, and roughly 10 cents of every project dollar is wasted on poor performance. None of this means T&M vendors are dishonest. It means the contract does nothing to stop drift.
Where fixed price goes wrong
Fixed price has its own failure mode: a price fixed on guesswork. When a vendor quotes a firm number after one sales call, the risk has not disappeared — it has been hidden as padding, or it resurfaces later as change-request battles and quiet quality cuts. A fixed price is only as trustworthy as the scoping work behind it. Warning signs: a quote before anyone has examined your systems and data, no written scope document signed by both sides, and no defined process for what counts as a change.
The hybrid that works: paid scoping, then a fixed price
The model that fixes both failure modes is a short, paid scoping phase followed by a firm quote. In our case that is a two-week scoping sprint: map the process, get access to the real systems and data, fix the scope, the success metrics, and the test plan. It ends with an architecture document, a fixed price, and a go/no-go decision — and if it is a no-go, you keep the document and have spent two weeks, not twelve. Delivery then runs in weekly demos of working software, which is exactly the iterative, small-batch approach that both PMI and the DORA research identify as the effective control on scope drift.
When time & materials is the right choice
T&M is honest and appropriate when “done” genuinely cannot be defined upfront: real research and development, an evolving product with a strong in-house product manager steering it week by week, or renting extra hands into a team you already run. In those cases a fixed price would just be fiction with a signature. The test is simple: if you cannot write down what finished looks like, do not sign a contract that pretends you can.
Questions to ask before signing either contract
Who writes the scope document, and do both sides sign it? What exactly counts as a change, and what does a change cost? What happens if the date slips — a penalty, free continuation, or nothing? Is payment tied to demonstrated working software or to calendar dates? And who owns the code, the infrastructure, and the documentation if you part ways mid-project? A vendor with clear answers to all five is safe under either model. A vendor without them is risky under both.
Isn't fixed price more expensive because vendors pad the risk?
Only when the price is fixed without real scoping — then padding is the only way a vendor can survive. A price fixed after a paid scoping sprint prices the actual work, not the uncertainty. If someone offers you a firm quote after a single call, assume padding, quality risk, or both.
What happens when requirements change mid-project?
There should be a written answer before you sign. Small swaps inside the agreed scope should be absorbed; genuinely new scope gets a quoted change order you can accept or decline. Weekly demos matter here too: drift gets caught in week five, not week eleven.
What if the vendor misses the deadline on a fixed-price contract?
Ask exactly this question before signing — the answer tells you how much the vendor believes its own plan. Our answer is written into the contract: live in 12 weeks on the signed scope, or we keep building at no cost until it ships.