Fixed-price vs time-and-materials is the first real argument in any software project, and it usually starts before anyone has agreed what the software actually does. One side wants a single number it can take to the board. The other wants room to discover what the build genuinely involves. Both positions are defensible, which is precisely why the conversation stalls.
The stakes are not academic. The same 3,200-hour build can be quoted at £180,000 under one commercial model and £240,000 under the other, and the cheaper quote is very often not the cheaper project. Signing the wrong contract shape does more than change the invoice: it changes who has an incentive to raise problems early, how change requests get priced, and whether the last month of your software development engagement is spent testing or arguing about scope. Getting fixed-price vs time-and-materials right at signature is worth more than any discount you will negotiate afterwards.
This guide treats fixed-price vs time-and-materials as a risk-allocation decision rather than a pricing one, because that is what it is. It covers how each model behaves in practice, what the cost-overrun research shows, the hidden costs on both sides, a project-type decision matrix, the hybrid models experienced buyers actually use, and the clauses that protect you whichever way you sign.
Table of contents
- Fixed-Price vs Time-and-Materials: The Difference That Actually Matters
- How a Fixed-Price Software Contract Really Works
- How a Time-and-Materials Contract Really Works
- Fixed-Price vs Time-and-Materials: What the Overrun Data Shows
- The Hidden Costs Neither Model Advertises
- Fixed-Price vs Time-and-Materials by Project Type
- Fixed-Price vs Time-and-Materials Risk Allocation
- Capped Time-and-Materials and the Hybrids That Work
- Choosing Between Fixed-Price vs Time-and-Materials: A Practical Framework
- Contract Clauses That Protect You in Either Model
- Frequently Asked Questions
- References and Further Reading
Fixed-Price vs Time-and-Materials: The Difference That Actually Matters
Most buyers frame this as a choice between a safe number and a risky one. That framing is wrong, and it leads to bad decisions in both directions.
It is a risk transfer, not a discount
A fixed price does not remove uncertainty from your project. It moves the financial consequences of that uncertainty onto the supplier, and the supplier charges you for accepting it. Time-and-materials leaves the same uncertainty on your side of the table and charges you less per hour because of it. The whole fixed-price vs time-and-materials debate collapses into one question: who is better placed to carry, and price, the unknowns?
Estimates never disappear — they get priced
Both models rest on the same estimate. A supplier quoting a fixed price has done the sums, added a contingency for the parts nobody can see yet, and given you the total. A supplier quoting time-and-materials has done the same sums and shown you the working. Nothing about fixed-price vs time-and-materials makes an inaccurate estimate accurate; it only decides who pays when the estimate is wrong.
The premium is real and it is measurable
In competitive UK mid-market tenders, the same scope quoted as fixed-price vs time-and-materials typically differs by 15% to 40%. That gap is the risk premium. On a well-specified integration it can be small enough to be worth paying for the certainty. On an exploratory product build it can be so large that you are effectively buying insurance against a risk you were going to manage anyway.
The head-to-head comparison
| Dimension | Fixed-price | Time-and-materials |
|---|---|---|
| Price certainty | High — one agreed total | Low — an estimate, not a commitment |
| Who carries scope risk | Supplier, within the specification | Buyer, in full |
| Risk premium paid | 15%–40% above hourly cost | None |
| Change handling | Formal change request, repriced | Reprioritise the backlog, no paperwork |
| Specification needed up front | Detailed and signed | Direction and priorities |
| Supplier incentive | Finish early, minimise effort | Keep the team busy |
| Buyer effort during delivery | Low day to day, high at acceptance | High throughout |
| Visibility of progress | Milestone reports | Continuous, if you look |
| Typical payment shape | Milestones tied to deliverables | Monthly against timesheets |
| Behaviour under pressure | Quality is the silent variable | Budget is the silent variable |
| Exit cost | Contested — what was delivered? | Clean — stop paying, stop working |
How a Fixed-Price Software Contract Really Works
The mechanics matter more than the principle, and this is where fixed-price vs time-and-materials stops being a philosophical argument. A fixed price is a promise about a document, not a promise about your business outcome.
Everything hinges on the specification
The supplier is contractually bound to deliver what the specification says, interpreted the way the specification reads — not the way you meant it. Every ambiguity becomes a negotiation later, and the party that wrote the ambiguity rarely wins. Producing a specification precise enough to underpin a fixed price is itself a project, usually a paid discovery phase of two to six weeks.
Contingency is inside the number you see
Suppliers do not gamble. They estimate the work, apply a loading for unknowns, and quote the total. On a clean, well-understood build that loading might be 15%. On a vague brief with an unfamiliar legacy system it can exceed 50%, and you will never see it itemised. You are paying that premium whether or not the risks materialise.
Change requests become the commercial engine
Once the price is fixed, the only route to more revenue is a change request. Some suppliers price change fairly. Others bid thin to win and recover margin through variations at a much higher effective rate. This is the single most common way a fixed-price vs time-and-materials decision reverses itself in hindsight: the fixed price wins the tender and loses the year.
Quality is the variable nobody writes down
When a fixed-price build runs late, the supplier has three levers: absorb the loss, argue scope, or reduce invisible quality. Test coverage, documentation, refactoring and accessibility are the first things to go, because none of them appear in an acceptance test. The cost lands on you eighteen months later as maintenance.
What good fixed-price delivery looks like
It is entirely achievable. It needs a specification both parties genuinely understand, acceptance criteria written before kick-off, milestones tied to demonstrable deliverables rather than dates, and a published change-request rate card. Suppliers who do this well tend to insist on discovery first — treat that insistence as a positive signal rather than an upsell.
How a Time-and-Materials Contract Really Works
Time-and-materials looks simpler and is harder to run. You are buying capability by the hour and taking responsibility for what it produces.
You buy capacity, not an outcome
The contract commits the supplier to provide named people with agreed skills at agreed rates, working on what you prioritise. It does not commit them to a finished product on a date. That distinction is easy to accept at signature and painful to discover at month five, which is why fixed-price vs time-and-materials arguments so often resurface halfway through delivery.
Rate cards hide more than they reveal
A £450 day rate is not comparable to a £650 day rate until you know seniority mix, whether project management and QA are billed separately, what the minimum billable increment is, and whether onboarding time is chargeable. Blended rates are convenient and hide the mix. Most fixed-price vs time-and-materials comparisons go wrong here rather than at the total. Always ask for the rate by role and the planned allocation by role.
Governance is the price of flexibility
Time-and-materials only produces good value when someone on your side owns the backlog, attends the demo, reads the burn rate weekly and is willing to say stop. Organisations without that capacity should not choose it. The agile methodologies that make the model work are governance disciplines, not a licence to leave requirements open.
Where time-and-materials quietly fails
It fails when nobody is watching. Budget drifts in small, individually reasonable increments; a two-week estimate becomes three because a dependency moved, and nobody adds up the pattern until the quarter closes. It also fails when the buyer treats an estimate as a commitment, which converts a collaborative model into a blame exercise.
The cap that makes it safer
Almost every mature buyer eventually adds a ceiling: work proceeds hourly, but the supplier must warn at 70% and cannot exceed the cap without written approval. That single clause removes most of the downside of time-and-materials while keeping the flexibility, and it is why the honest answer to fixed-price vs time-and-materials is often “neither, quite”.
Fixed-Price vs Time-and-Materials: What the Overrun Data Shows
Both camps quote research at each other. The research is less flattering to both than either side admits.
Overruns cluster around scope, not coding
Across large IT programmes, the dominant causes of budget overrun are changes to requirements after sign-off, underestimated integration work, and unclear ownership of decisions — not slow developers. That matters for fixed-price vs time-and-materials because a fixed price only protects you from the causes that sit inside the specification, and most of them sit outside it.
Where software budget overruns actually come from
Indicative distribution for mid-market bespoke builds, consistent with the published programme research listed in the references.
The tail risk is the real story
Average overruns are unremarkable; the distribution is not. A minority of projects overrun catastrophically, and those outliers dominate the total loss. A fixed price genuinely protects you against that tail — that is its strongest argument, and it is worth a premium if a catastrophic overrun would threaten the business rather than merely annoy the board.
Neither model prevents an overrun
A fixed-price contract does not stop a project failing; it converts failure into a dispute. If the supplier cannot deliver profitably, you get slipped dates, thin quality and a legal argument, not a refund. Any honest fixed-price vs time-and-materials comparison has to price the probability of that dispute alongside the certainty of the number.
What actually correlates with success
The variables that predict on-budget delivery are the same in both models: a small number of clearly owned decisions, short feedback loops, an engaged product owner, and technical discovery completed before the expensive phase starts. Fixed-price vs time-and-materials is a second-order factor next to any of those, which is a genuinely uncomfortable conclusion for everyone selling one.
The quoted price is not the cost of the project. Each side of fixed-price vs time-and-materials carries overheads that appear in a different budget line, and neither appears in the proposal.
Fixed-price: the change-request multiplier
Work done inside the original price is charged at the quoted blended rate. Work done through a variation is charged at whatever the change-control clause permits — typically a materially higher effective rate, plus the internal cost of specifying and approving each one. Three or four variations on a mid-sized build routinely erase the certainty you paid for.
Effective cost of the same feature, by route (indexed: in-scope work = 100)
The ordering here is stable across engagements even where the exact multiples differ; the cost of a decision rises with how late it is made.
Time-and-materials: the management overhead
Running the model properly costs perhaps 0.2 to 0.5 of a full-time internal role for the duration — backlog grooming, demo attendance, burn-rate review, prioritisation calls. That is a real cost that never appears on the supplier’s invoice, and leaving it out of a fixed-price vs time-and-materials calculation is how organisations conclude that hourly billing “went over budget” when in fact nobody was steering.
Both: the cost of the wrong specification
The most expensive line item in either model is building the right thing badly or the wrong thing well. Discovery, prototyping and user testing look like delay and are the cheapest risk reduction available. This is where technology consulting earns its fee, and it is entirely independent of how you eventually contract the build.
And the cost of comparing badly
Buyers routinely compare a fixed price that includes project management, QA, environments and three months of warranty against an hourly rate that includes none of them. Normalise the two quotes onto the same inclusions before drawing any conclusion — otherwise the fixed-price vs time-and-materials comparison you are running is arithmetic on incompatible numbers.
Fixed-Price vs Time-and-Materials by Project Type
The right answer changes with the work, which is why fixed-price vs time-and-materials cannot be settled as a company-wide policy. This matrix reflects what tends to hold in practice rather than what either sales pitch claims.
Well-defined integrations and migrations
When the endpoints are documented, the data model is known and success is objectively testable, fixed-price is usually correct. The uncertainty is genuinely low, so the risk premium is small, and you get a clean commercial boundary around a bounded piece of work.
Discovery-heavy products and MVPs
When you are learning what the product should be, a fixed price forces you to guess the specification and then pay to change every guess. Iterative work belongs on time-and-materials or capped time-and-materials — which is why most MVP development engagements are structured that way.
Long-running platforms and product teams
Once software becomes a permanent capability rather than a project, per-project pricing stops making sense and fixed-price vs time-and-materials stops being the relevant question at all. A stable squad on time-and-materials, funded quarterly against a roadmap, is the normal shape — and where DevOps practices make throughput visible enough for the model to be governed honestly.
| Project type | Recommended model | Why |
|---|---|---|
| Documented API integration | Fixed-price | Low uncertainty, objectively testable, small premium |
| Data migration, known source | Fixed-price | Bounded scope; profile the data during discovery |
| Website or brochure rebuild | Fixed-price | Mature patterns, visual acceptance criteria |
| MVP or new product | Capped T&M | Scope must flex on evidence; cap protects the budget |
| Legacy modernisation | T&M after paid discovery | Unknowable until you open the code |
| Ongoing product squad | T&M, quarterly funding | Continuous roadmap, no fixed end state |
| Regulatory or compliance deadline | Fixed-price per phase | Date is non-negotiable; buy certainty in slices |
| Research or AI feasibility work | T&M, time-boxed | Outcome genuinely unknown; limit the spend, not the scope |
| Small, well-understood enhancement | Fixed-price | Cheaper to quote than to govern |
Fixed-Price vs Time-and-Materials Risk Allocation
Read the two models as a risk register and the decision becomes considerably less philosophical.
Reading a risk register before you sign
List the ten things most likely to go wrong on your specific project, then write down who pays if each one happens under each model. Most buyers discover that the risks they were worried about — a shifting market, an internal reorganisation, a slow third party — sit on their side of the line in both contracts. That exercise settles a fixed-price vs time-and-materials debate faster than any spreadsheet.
Why suppliers price risk higher than you do
A supplier carrying your scope risk has less information than you about your own organisation and cannot control your decision-making speed. Rational pricing of a risk you cannot control is expensive pricing. Where you are demonstrably the better manager of a risk, paying someone else to carry it is poor cost optimisation.
Security and compliance obligations both models share
Neither model changes your regulatory position. Data protection duties, cybersecurity obligations, penetration testing, secure development practices and breach notification all remain yours regardless of how the fixed-price vs time-and-materials question was answered, and they must be written into the agreement explicitly. A fixed price that quietly excludes security testing is not a cheaper project — it is an unfinished one.
The clauses that move risk back
Warranty length, defect-fix obligations, liability caps, service credits and acceptance rights all shift risk after signature. A generous fixed price with a two-week warranty and a liability cap of 10% of fees may transfer far less risk than a time-and-materials contract with a six-month defect obligation. Any fixed-price vs time-and-materials judgement made from the headline numbers alone is therefore incomplete — read those clauses first.
| Risk | Under fixed-price | Under time-and-materials |
|---|---|---|
| Effort exceeds the estimate | Supplier | Buyer |
| Requirements change | Buyer (via change request) | Buyer (via reprioritisation) |
| Third party or dependency slips | Usually buyer, check the clause | Buyer |
| Technical approach proves wrong | Supplier | Buyer |
| Slow internal decisions | Buyer, and it voids delay claims | Buyer, directly and immediately |
| Quality below expectation | Buyer, unless acceptance criteria are tight | Buyer, unless governance is active |
| Key developer leaves | Supplier | Shared — you pay the ramp-up |
| Project cancelled early | Contested; termination clause governs | Buyer pays work done, then stops |
| Post-launch defects | Supplier during warranty | Buyer buys the fix time |
How delivery risk splits between the parties
Deep violet = risk carried by the supplier. Lavender = risk carried by the buyer. The premium you pay tracks the violet share.
Capped Time-and-Materials and the Hybrids That Work
Very few experienced buyers choose either pure model. The useful answer to fixed-price vs time-and-materials is almost always a structure that borrows from both.
Capped time-and-materials
Work is billed hourly, but a contractual ceiling applies, with a mandatory warning at an agreed percentage of the cap. You keep the flexibility to reprioritise and the supplier keeps the incentive to be efficient, because exceeding the cap needs your written agreement. It resolves fixed-price vs time-and-materials by taking the certainty from one and the adaptability from the other, and it is the default recommendation for most mid-market builds.
Fixed-price per phase or per sprint
Break the work into slices and fix the price of each one as you reach it, using what the previous slice taught you. Certainty improves with every phase because the estimate is made closer to the work. It suits regulatory deadlines and boards that need a number, without demanding a complete specification on day one.
Target cost with a shared band
Both parties agree a target, and variance inside an agreed band is shared — for example 50/50 up to 15% either way. It aligns incentives better than either pure model, and it requires open-book cost transparency and genuine trust, so it works with a supplier you already know rather than a new one.
Fixed budget, flexible scope
The commercial commitment is the money and the date; the variable is what gets built. You prioritise ruthlessly and ship the most valuable subset within the envelope. This is often the most honest structure available and it is culturally the hardest to sell internally, because someone must accept that the feature list is negotiable.
Discovery fixed, build flexible
The simplest hybrid of all: buy a fixed-price discovery phase whose deliverable is a specification and an estimate, then decide the build model with real information. It costs a few weeks and typically pays for itself by removing the guesswork that makes both a fixed price expensive and a time-and-materials budget unpredictable. It also converts fixed-price vs time-and-materials from a bet into an informed choice.
Choosing Between Fixed-Price vs Time-and-Materials: A Practical Framework
Work through this in order. The first question that returns a clear answer usually settles it.
Score your certainty first
Can you write acceptance criteria today that you would be content to be held to in twelve months? If yes, fixed-price is available to you. If you cannot — because the market, the users or the legacy system will teach you something — then a fixed price is a fixed price for the wrong thing, however comforting the number looks. Certainty of scope, not appetite for certainty, is what decides fixed-price vs time-and-materials.
Then score your governance capacity
Do you have a named person with authority to prioritise, at least half a day a week, for the whole engagement? If not, time-and-materials will disappoint you regardless of the supplier’s quality. Honest self-assessment here resolves more fixed-price vs time-and-materials disputes than any commercial analysis.
Then decide what you are optimising for
Certainty, speed, total cost and quality cannot all be maximised at once. Fixed-price optimises certainty. Time-and-materials optimises adaptability and usually total cost. Fixed budget with flexible scope optimises time to value. Name your priority explicitly, because an unnamed priority defaults to whichever one the supplier prefers.
A five-question test
Ask: is the scope objectively testable; will requirements change; can we cope with a 40% overrun; do we have a product owner; and is this a project or a permanent capability? Three or more answers favouring flexibility means capped time-and-materials. Three or more favouring certainty means fixed-price per phase. Five questions settle fixed-price vs time-and-materials more reliably than a month of proposals.
Then negotiate the shape, not just the rate
Once the model is chosen, the value is in the clauses — caps, warnings, change rates, warranty, acceptance. Suppliers expect to negotiate these and buyers rarely try. A well-structured outsourcing strategy spends its energy there rather than grinding two percent off a day rate.
Contract Clauses That Protect You in Either Model
Whichever way the fixed-price vs time-and-materials question resolves, the same handful of clauses determine whether the contract helps you when things go wrong.
Acceptance criteria written before kick-off
Define what “done” means for each deliverable, in testable language, before work starts. Vague acceptance language is the root of most fixed-price disputes and most time-and-materials disappointments, so it is the one clause that matters equally under fixed-price vs time-and-materials. If a criterion cannot be tested by someone who was not in the room, rewrite it.
A change-control procedure with a published rate
State how a change is raised, who prices it, how long the buyer has to decide, and at what rate the work will be charged. A rate card agreed at signature costs nothing and removes the single largest source of mid-project friction under a fixed price.
Key-person and continuity clauses
Name the people you are buying, require notice and equivalent replacement, and require a handover overlap. Under time-and-materials, also specify that ramp-up for a supplier-initiated change is not chargeable. Team churn is one of the most common and least-discussed sources of quiet overrun.
Intellectual property, escrow and exit
Confirm that IP transfers on payment rather than on final acceptance, that source code and infrastructure-as-code are delivered continuously into a repository you own, and that an exit plan exists. Effective vendor management assumes the relationship will end one day, ideally amicably.
Reporting you will actually read
Require a short weekly report: work completed, work planned, spend to date, spend forecast, risks. One page, every week, in both models. Most overruns are visible in this document for six weeks before anyone acts on them.
Frequently Asked Questions
Is fixed-price always more expensive than time-and-materials?
No — but it usually is, for the same delivered scope. You pay a 15% to 40% risk premium for certainty. Fixed-price ends up cheaper when it forces the discipline that prevents scope drift, or when a time-and-materials engagement runs ungoverned. The fixed-price vs time-and-materials cost comparison depends more on your governance than on the rate card.
Can we start fixed-price and switch to time-and-materials?
Yes, and the reverse is common too. A fixed-price discovery followed by a capped time-and-materials build is the most widely used sequence in UK mid-market software work, and it is the reason fixed-price vs time-and-materials rarely has to be answered once and for all. Agree the transition mechanism in the original contract so the switch is a planned step rather than a renegotiation under pressure.
What is a reasonable cap on a capped T&M contract?
Usually the estimate plus a 15% to 20% contingency, with a written warning obligation at 70% and 90% of the cap. Set the cap on the phase rather than the whole programme so both parties recalibrate regularly, and require that unused contingency is simply not billed.
How do we compare quotes fairly across the two models?
Normalise inclusions first: project management, QA, design, environments, warranty and support must be in or out of both. Then convert the time-and-materials estimate to a total at the mid-point of its range, add your own internal governance cost, and compare that against the fixed price. Any fixed-price vs time-and-materials comparison that skips this step is misleading.
Does agile delivery rule out fixed-price?
Not entirely, though it constrains it. Fixing the price of a defined phase or sprint is compatible with agile delivery; fixing the price of a twelve-month backlog is not, because the model exists to let evidence change the plan. Fixed budget with flexible scope is the usual reconciliation.
Which model do suppliers actually prefer?
Good suppliers prefer whichever model matches the uncertainty of the work, and they will tell you so — a supplier with a fixed view on fixed-price vs time-and-materials regardless of the project is selling a preference, not advice. Be cautious of any supplier that offers a firm fixed price for an ill-defined build without insisting on discovery first: that price contains either a very large contingency or a plan to recover margin through variations.
What should we do if we genuinely cannot decide?
Buy a paid discovery phase at a fixed price. It is the one answer to fixed-price vs time-and-materials that is almost never wrong: it is small, bounded, produces a specification and an estimate, and leaves you free to choose the build model with real information rather than assumptions.
References and Further Reading
The sources below cover the contractual definitions, the estimation research and the overrun evidence behind the figures used above. They are the primary references worth reading before any fixed-price vs time-and-materials decision reaches a board paper.
Fixed-price contract — definition, variants and public-sector usage (Wikipedia)
Time and materials — how the model is structured and where it is used (Wikipedia)
Cost overrun — evidence across large projects and programmes (Wikipedia)
Scope creep — causes and controls (Wikipedia)
Principles behind the Agile Manifesto — the source text on responding to change
GOV.UK Service Manual — agile delivery guidance for government projects
Software development effort estimation — why estimates behave the way they do
Reference class forecasting — the corrective for optimistic estimates
