Fixed price works when the scope is genuinely known: a defined MVP, a migration with a clear end state, a website rebuild. Time and materials works when you are discovering the product as you build. The model most senior UK buyers actually use in 2026 is a hybrid: a fixed-price discovery phase of 2 to 4 weeks, then capped time and materials sprints with a named team and exit rights at every sprint boundary. The pricing model matters less than three contract terms: who owns the IP, what happens when scope changes, and how easily you can leave. This guide covers all of it in plain English.
Every software project starts with the same commercial question: do we agree a price for the whole thing, or pay for the time it takes? Agencies have strong opinions, usually the one that suits their cash flow. Buyers deserve a straight answer, so here it is.
What each model actually is
Fixed price: you agree scope, timeline and a single number. The vendor carries the delivery risk. If it takes longer, they absorb it. In exchange, the price includes a risk premium, typically 20 to 40 percent above the same work billed by time, and every change to scope becomes a formal negotiation.
Time and materials (T&M): you pay day rates for the team, monthly in arrears. You carry the delivery risk, and in exchange you get flexibility: priorities can change every sprint without a change request. The bill is only acceptable if you actively steer the backlog.
| Dimension | Fixed price | Time & materials |
|---|---|---|
| Who carries risk | Vendor (priced in at 20 to 40 percent premium) | Buyer (needs active steering) |
| Scope changes | Formal change requests, friction, delay | Reprioritise next sprint, no ceremony |
| Quality incentive | Vendor incentivised to cut corners to protect margin | Vendor incentivised to keep the work going |
| Budget certainty | High, if scope truly holds | Low without caps, high with caps |
| Best for | Defined MVPs, rebuilds, migrations, integrations | Evolving products, R&D, long-term squads |
When fixed price is the right call
Fixed price is honest when the unknowns are small. Four signs you can fix the price safely:
- The end state is describable in one document. Screens, integrations, data model, acceptance criteria. If you cannot write it down, nobody can price it.
- It has been done before. A Shopify build, a standard API integration, a lift-and-shift migration. Precedent means predictable effort.
- The deadline matters more than the wish list. Fixed price forces scope discipline, which is a feature when a launch date is immovable.
- You will not be available to steer weekly. If nobody on your side can attend sprint reviews, T&M drifts. Fixed price protects an absent buyer better.
Watch for the lowball: a fixed bid dramatically cheaper than the rest is not a bargain, it is a plan to recover margin through change requests. The cheapest credible bid and the most expensive one should sit within roughly 40 percent of each other. An outlier below that range is a red flag, not a saving.
When time and materials is safer
T&M is honest when discovery is part of the work. If the product will change as users react, fixing the price just means paying the risk premium and then paying for change requests on top. Choose T&M when:
- You are building something new and expect to learn from each release.
- The backlog will be reprioritised monthly or faster.
- You want a standing squad that behaves like your own team.
- You have someone senior who will review progress every sprint and cut work that is not paying back.
The discipline that makes T&M safe is a cap. A monthly or quarterly ceiling, agreed in writing, with a rule that the vendor flags at 70 percent burn. Uncapped T&M with a passive buyer is how £80k projects become £300k projects. We have seen it. More than once.
The hybrid model senior buyers actually use
In 2026 the default commercial shape we recommend, and the one our strongest vendors prefer too, is neither pure model. It is three stages:
- Stage 1: fixed-price discovery, 2 to 4 weeks. A small senior team produces the spec, architecture, backlog and a costed delivery plan. Typical UK cost: £5,000 to £25,000. You own everything produced, and you can take it to any vendor.
- Stage 2: capped T&M delivery in sprints. Named team, agreed day rates, monthly cap, demo every sprint, exit right at every sprint boundary with 2 weeks notice.
- Stage 3: fixed-price hardening and handover. Documentation, tests, deployment runbooks, a warranty period of 30 to 90 days for defects.
This shape gives you budget certainty where certainty is possible and flexibility where it is not. The discovery output also does something quietly important: it makes the delivery competitive. A vendor who knows you hold a complete spec behaves differently from one who knows you are locked in.
Five contract questions that matter more than the pricing model
- Who owns the IP, and when? The right answer: you, on payment, including work in progress. Not on final completion.
- What exactly triggers a change request? Get examples in the contract. Vague scope language is where fixed-price projects go to die.
- Who are the named people, and what happens if they change? You are buying the team that impressed you, not the bench. Substitution needs your approval.
- What does leaving look like? Notice period, handover obligations, access to repositories and environments from day one. If leaving is painful, pricing power is theirs, forever.
- What is warranted after go-live? A defect warranty of at least 30 days, separate from any support retainer they would like to sell you.
Quick decision guide
- IF scope is defined and precedented Fixed price, with change-trigger examples written into the contract.
- IF the product will evolve as you learn Capped T&M with a named team and sprint-boundary exit rights.
- IF you are not sure what you need yet Buy a fixed-price discovery first. £5k to £25k buys certainty and negotiating power.
- IF a bid is far below every other bid Treat it as a change-request business model. Walk away or re-scope it.
The honest summary
Pricing models do not rescue bad scoping, and they do not sink good teams. Fix the price when the work is knowable. Pay for time when the work is discovery. Cap everything. Own your IP from day one. And remember that the best predictor of a software project's outcome is not the commercial model, it is whether senior people on both sides look at working software every two weeks.
If you would rather have someone run that whole process for you, that is literally what we do: scope the brief, source two or three vetted UK teams, put their quotes side by side in plain English, and manage delivery to handover. One brief in, one outcome out.
Scoping a software project?
Send us the brief. A human replies within 30 minutes, day or night, with next steps and a straight recommendation on the right commercial model for your case.
Start a brief →