The short answer
What this means for you
Fixed price pays an agreed total for an agreed scope, time and materials pays for hours worked at agreed rates, and fixed capacity pays for a team's time over a period. Choose fixed price when the scope is small, stable, and testable. Choose time and materials, with a cap or approval points, when you expect to learn and change course. Choose fixed capacity for ongoing product work. If scope is uncertain, buy a short discovery first. Whatever the model, define change control, acceptance, ownership, warranty, and handover.
- Fixed price moves effort risk to the supplier, so expect a buffer in the price and a formal change process.
- Time and materials keeps you flexible but puts effort risk on you, so add caps, reporting, and frequent working demonstrations.
- Fixed capacity buys a team's time for a period: you keep the scope risk and decide the priorities, and the supplier carries staffing continuity.
- In the US, UK, Canada, and Australia, an independent supplier can keep copyright in work you commission unless a signed written agreement deals with ownership.
- Tie payments to accepted, working deliverables and agree a warranty period, a transition-out plan, and who holds every account.
Should you choose a fixed price or time and materials contract for software development?
Fixed price: you pay an agreed total for an agreed scope. Time and materials (also written time and material, or T&M): you pay for the hours worked on your software at agreed rates. Fixed capacity: you pay an agreed amount for a team's time over a period and decide what it works on.
Choose fixed price when the scope is small, stable, and specific enough to test, such as a marketing website with agreed pages and content. Choose time and materials, ideally with a spending cap and regular approval points, when you are building a web app or MVP and expect to change the plan as you learn. If you cannot yet write testable acceptance criteria, buy a short, separately priced discovery phase first and agree the build contract afterwards.
In short, which is better depends on how much of your scope you can already write as testable statements. Fixed price wins when you can, time and materials wins when you cannot, and fixed capacity wins once the product is live and the backlog keeps moving.
Our view: The pricing model is only one clause. Most disputes come from what the contract leaves vague: how changes are approved, what counts as done, who owns the code and accounts, what happens to bugs found after launch, and how you leave if the relationship ends. This guide covers each of those in turn and ends with a checklist you can use before you sign.
This guide is general information for business owners and founders in the United States, United Kingdom, Canada, and Australia. It is not legal advice. Contract law and intellectual property law differ by country, state, and province, so have a qualified lawyer review any agreement before you sign it.
What are the main pricing models for a software development contract?
Guidance for the UK government's Digital Outcomes and Specialists framework, withdrawn on 31 December 2025 as out of date, described three payment approaches. It said a fixed price approach can be more expensive because the supplier owns most of the risk, and that time and materials means the buyer owns more of the risk. It described capped time and materials as working like time and materials but with a limit on what the buyer pays, and said that if the limit is reached before the work is finished, the supplier has to complete the work at its own cost.[1]
The closest official definitions come from US federal procurement. Under FAR 16.202-2, a firm-fixed-price contract is suitable when fair and reasonable prices can be set at the outset, including where performance uncertainties can be identified and their cost impact reasonably estimated. This is a US federal definition for government buying, not private-contract law, but it states the idea plainly: fixed price works when the risks can be priced.[17]
FAR 16.601 describes a time-and-materials contract as one that pays fixed hourly rates for labour and actual cost for materials. It requires a ceiling price "that the contractor exceeds at its own risk", and allows this contract type only when it is not possible at the time of placing the contract to estimate accurately the extent or duration of the work. That is the same logic as capped time and materials in private work: flexibility for unknown scope, with a limit on what you can be asked to pay.[16]
For fixed capacity, the nearest official analogue is the FAR's firm-fixed-price, level-of-effort term contract. FAR 16.207-1 says the contractor provides a specified level of effort, over a stated period of time, on work that can be stated only in general terms, and the buyer pays a fixed dollar amount. Again, this is a US federal definition, not a rule for private contracts.[15]
Our view: How risk moves between buyer and supplier: under fixed price the supplier carries effort risk and you carry the risk that the written scope is wrong. Under time and materials you carry effort risk, and the supplier carries the risk of poor estimates only to the extent a cap or goodwill makes them. Under fixed capacity you carry scope and priority risk, and the supplier carries staffing continuity: the named people, their seniority, and replacement when someone leaves.
The UK government's Digital, Data and Technology Playbook states that traditional pricing models such as fixed price or time and materials for the whole project are not optimal for agile work, and that an outcome-based contract built on collaboration is likely to be more appropriate. It also says fixed pricing may be appropriate where the scope of a project is certain.[2]
Private contracts use the word cap loosely. In some, the supplier must finish the agreed scope within the cap. In others, the cap is only a point where work stops until you approve more budget. Ask which one a quote means and make the contract say it plainly.
| Model | How you pay | Who carries effort risk | Fits best when | Watch for |
|---|---|---|---|---|
| Fixed price | An agreed total for an agreed scope, usually split across milestones. | Mostly the supplier, within the written scope. | Scope is small, stable, and written as testable acceptance criteria, such as a defined website build. | A risk buffer in the price, narrow scope wording, and every change becoming a priced change request. |
| Time and materials | Hours or days worked at agreed rates, usually invoiced weekly or monthly. | Mostly you, the buyer. | Requirements will change as you learn, such as an MVP or ongoing product work. | Open-ended spend without reporting, budget alerts, or regular working demonstrations. |
| Capped time and materials | Time and materials up to an agreed ceiling for an agreed scope or period. | Shared. Depends on what the contract says happens at the cap. | You want flexibility but need a firm budget limit for a phase. | Whether the supplier must finish at its own cost or simply stops work when the cap is reached. |
| Phased: discovery, then build | A small, separately priced discovery, then a new price or budget for the build. | Limited to one phase at a time. | The problem is clear but the solution, integrations, or data are not. | Discovery outputs you do not own, or a build price that ignores what discovery found. |
| Fixed capacity (monthly team) | An agreed fee per sprint or month for named roles. | You carry scope risk. The supplier carries staffing continuity. | Ongoing product work with a backlog, such as post-launch improvements. | What happens to unused capacity, and how output is reported. |
What is a fixed capacity model, and when does it beat fixed price?
Our view: A fixed capacity model buys a team rather than a deliverable: you agree a fee per sprint or month for named roles, such as two developers and a part-time designer, and you decide what the team works on from a backlog. It beats fixed price when the work cannot be stated as a finished list, which is typical of a product that is live and changing every week. It beats open-ended time and materials when you want a predictable monthly cost.
Our view: The trade-off is that fixed capacity prices time, not outcomes. If priorities are vague, you still pay for the team. The contract should therefore make output visible through a regular demonstration and a short report of what shipped against what was planned.
- Roles and named seniority: who is on the team, at what level, and what share of their time you get.
- Replacement rules: how fast a departing person is replaced, and who pays for the handover learning time.
- Planning cadence: how often priorities are set, and who on your side has the final say.
- Reporting format: a working demonstration each sprint and a short written summary of work done.
- Unused capacity: whether it carries over, is credited, or is lost, and what counts as unused.
- Notice and scale: how much notice either side gives to change team size or end the agreement.
Which contract model fits a website, a web app, or an MVP?
Our view: Match the model to how much you already know. The more of your scope you can write as testable statements today, the more a fixed price makes sense. The more you expect user feedback to change the product, the more you need a model that prices change cheaply.
The Manifesto for Agile Software Development values customer collaboration over contract negotiation and responding to change over following a plan. Its principles welcome changing requirements, even late in development, and say working software is the primary measure of progress.[3][4]
The GOV.UK Service Manual says you should not start building your service in discovery, that there is no set time period but around 4 to 8 weeks is typical, and that it is not a failure to stop at the end of discovery if research shows that is the best thing to do.[5]
For a private business a much shorter discovery is often enough. What matters is the output: a written task list, integration review, risk list, acceptance criteria, and a revised estimate that you own and could take to another supplier.
Not sure how certain your scope is? The free software project estimator gives an effort range for your scope and suggests a contract model to start the conversation with suppliers.
- Brochure or marketing website with agreed pages, content, and integrations: fixed price usually works, provided content delivery dates and design rounds are written down.
- Website migration or redesign with known pages and redirects: fixed price for the agreed inventory, with a rate card for pages or redirects added later.
- MVP for a new product: time and materials or capped time and materials in short phases, so you can cut or reorder features after each demonstration.
- Web app with complex integrations, payments, or regulated data: a paid discovery phase first, then a fixed price or capped phase for the build once unknowns are resolved.
- Ongoing product work after launch: fixed capacity, a monthly team agreement, or time and materials, with clear priorities and reporting.
- Rescue of an existing codebase (see the software project rescue service): a short, separately priced assessment before any commitment to fix, rebuild, or replace.
How should change requests work in a software development contract?
Every software project changes. The contract should make change cheap to discuss and clear to approve, so neither side starts unapproved work or argues about it at invoice time.
Our view: Under fixed price, a change request usually adds cost or time, so the process must be quick and the pricing predictable. Under time and materials, a change mostly shifts priorities within the same budget, so the important control is who can reorder work and how you see the effect on the remaining budget.
- Who can request a change, and who on your side can approve one in writing.
- How quickly the supplier must respond with an impact on cost, schedule, and other features.
- Which rate card applies to priced changes, so you are not negotiating each one from scratch.
- Whether small changes under an agreed threshold can be swapped for work of similar size without a new price.
- That no change is billable until it has written approval, and that approved changes are recorded in one log.
- How removing a feature reduces the price under fixed price, not only how adding one increases it.
How should milestones, acceptance, and payments work?
The US Digital Services Playbook, written for government buyers, recommends contracts that request frequent deliverables rather than multi-month milestones, hold vendors accountable to deliverables, and give the buying team flexibility to adjust feature priorities and delivery schedules. It also recommends budgeting for research, discovery, and prototyping.[6]
Write each milestone as working software you can try, not as a document or a percentage complete. An acceptance criterion should be a statement someone can check, such as a customer can request a booking on a phone and receives a confirmation email, rather than booking flow complete.
Agree how acceptance happens: where you test, how long you have to review, what counts as a defect versus a new request, and what happens if you do not respond in time. Some contracts treat silence after the review window as acceptance, so if yours does, make sure the window is realistic for your team.
The UK Digital, Data and Technology Playbook advises selecting payment milestones so that the delivery of non-functional requirements is included, not only visible features.[2]
Tie each fixed price payment to an accepted milestone, not to a calendar date or a percentage the supplier reports. If the supplier asks for a deposit at signing to reserve people for your project, keep it to a small share of the total. Keep a meaningful share for final acceptance and handover.
For time and materials, ask for timesheets or a weekly report of hours against planned work, an alert before any agreed budget threshold, and the right to pause work. Check the currency, tax treatment, invoice terms, and whether third-party costs such as hosting or paid APIs are passed through at cost or with a margin.
- Each milestone names the user tasks it delivers and the checks they must pass.
- Non-functional requirements are included: accessibility, performance, security, browser and device coverage, and backups.
- You review on a test environment with realistic sample data, not screenshots.
- Defects are graded by severity, and only agreed severities block acceptance.
- The review period, the number of fix rounds, and the sign-off method are written down.
Who owns the code, IP, and accounts when an agency builds your software?
This section is general information about how copyright ownership works by default in four countries. It is not legal advice. Ownership of software can also involve patents, trade secrets, licences, and moral rights, so ask a lawyer in your jurisdiction to review your IP clause.
United States: the US Copyright Office explains that a work is made for hire in two situations: when an employee creates it as part of regular duties, or when it is specially ordered or commissioned, falls within one of nine listed categories, and both parties expressly agree in a signed written instrument that it is a work made for hire. The nine categories are a contribution to a collective work, part of a motion picture or other audiovisual work, a translation, a supplementary work, a compilation, an instructional text, a test, answer material for a test, and an atlas.[7]
Our view: Custom software is not one of those named categories, so a work made for hire clause on its own may not give you ownership of code written by an independent agency. US copyright law separately states that a transfer of copyright ownership is not valid unless it is in writing and signed by the owner of the rights conveyed or the owner's authorized agent. This is one reason agency contracts often include both a work made for hire clause and a written assignment of copyright.[7][8]
United Kingdom: GOV.UK says that if you are self-employed, you usually own the intellectual property even if your work was commissioned by someone else, unless your contract gives them the rights. It also says that to sell or transfer copyright you need to write and sign a document, sometimes called an assignment.[9][10]
Canada: the Copyright Act says the author of a work is the first owner of the copyright, that an employer is the first owner of work made by an employee in the course of employment unless there is an agreement to the contrary, and that no assignment or grant is valid unless it is in writing and signed by the owner of the right. The Canadian Intellectual Property Office notes that moral rights cannot be assigned or transferred but can be waived.[11][12]
Australia: IP Australia says IP created by a contractor is the property of the contractor unless otherwise stated in the contract, and recommends a written agreement before work begins. business.gov.au says that if your business wants to own the IP, the contract must clearly say this.[13][14]
Our view: The practical point is the same in all four countries: do not assume that paying for software makes you its owner. Put ownership in a signed written agreement, say when it transfers, and deal separately with the agency's pre-existing tools and with open-source components, which stay under their own licences.
The US Digital Services Playbook also recommends that contracts specify that software and data generated by third parties remain under the buyer's control. Apply the same idea to accounts: the domain, hosting, code repository, app store, analytics, email delivery, and payment accounts should be registered to your business, with the agency given access rather than ownership.[6]
What warranty and bug-fix period should a software contract include?
The US Digital Services Playbook recommends that a contract specifies a warranty period in which the vendor fixes defects found by the public at no additional cost to the buyer.[6]
Define what the warranty covers before you sign. A useful warranty covers defects against the accepted acceptance criteria for a stated period after launch, with response times by severity. It usually excludes new features, changes you or another supplier make to the code, and failures in third-party services the supplier does not control.
Keep warranty fixes separate from support and maintenance. Security patches, dependency updates, hosting changes, and small improvements after launch are ongoing work, so price them as a separate support agreement with its own scope and response times.
What should the contract say about termination and handover?
The US Digital Services Playbook recommends that contracts include a transition of services period and a transition-out plan. The UK Digital, Data and Technology Playbook says contracts should include a requirement to develop an exit plan and obligations on the supplier to supply data and information back at the end of the contract, and recommends breakpoints at the end of each stage of a pilot.[6][2]
Agree how either side can end the contract, the notice period, and what you pay for on exit: usually accepted work plus work in progress up to the termination date. The contract should require the supplier to hand over source code, documentation, credentials, and data in a usable form on termination for any reason, once undisputed invoices are paid.
Our view: Phased contracts make exit easier. If each phase ends with working software, documentation, and a decision point, you can stop or change supplier without a dispute about how much of a large fixed scope was finished.
- Source code in a repository your business owns, at the revision running in production.
- Setup, build, and release instructions that a new developer has tested.
- An inventory of accounts, environments, and third-party services, with billing owners.
- Database exports, file storage, and backups, with a tested restore.
- A list of known defects, open work, and the agreed handover support period.
What are the red flags in a software development contract or quote?
A quote shows how a supplier will behave when things change. Read it for what it leaves out as well as for the total.
A quote far below the others is a signal to ask questions, not to reject or accept it. Our guide on how to compare software development quotes covers what to ask a low bidder, including the GOV.UK guidance on abnormally low tenders, and shows a line-by-line method for putting two or more quotes on the same footing.
- A fixed price for a vague scope, with no acceptance criteria and no written assumptions.
- No change process, or a change rate far higher than the rate used to price the original scope.
- IP that transfers only on final payment with no handover of work in progress if you end the contract early.
- Hosting, domain, repository, or app store accounts registered to the supplier rather than to your business.
- Payments tied to calendar dates or reported percentages instead of accepted, working deliverables.
- No warranty period, or a warranty that excludes almost every likely defect.
- Testing, accessibility, security, or deployment missing from the quote or listed as optional extras.
- Time and materials with no reporting, no budget alert, and no right to pause.
- A large upfront payment before any discovery, design, or working software exists.
- Pressure to sign quickly, or reluctance to let your lawyer review the agreement.
Software development contract checklist: what should the agreement include?
Use this checklist to compare agreements from different suppliers side by side. It is a starting point for your own review and your lawyer's, not a substitute for legal advice.
- Scope: the user tasks, integrations, platforms, and exclusions, with assumptions that would change the price.
- Pricing model: fixed price, time and materials, capped, or phased, and exactly what happens at any cap.
- Change control: who requests and approves changes, response times, rate card, and a single change log.
- Milestones and acceptance: working deliverables, testable criteria, review periods, and defect severity rules.
- Payment schedule: deposit, milestone payments tied to acceptance, currency, tax, invoice terms, and pass-through costs.
- IP and ownership: a signed written assignment or equivalent clause for your jurisdiction, pre-existing tools, open-source licences, and moral rights where relevant.
- Accounts and data: domains, hosting, repositories, app stores, and payment accounts held by your business.
- Warranty and support: the defect warranty period, what it covers, and a separate support agreement.
- Confidentiality, data protection, and security responsibilities for personal or regulated data.
- Termination and exit: notice, payment on exit, handover materials, transition support, and data return.
Common questions about software development contracts
Which is better, fixed price or T&M? Neither is better in general. Fixed price is better when the scope is small, stable, and testable, because the supplier carries the effort risk. Time and materials is better when you expect to learn and change course, because you pay only for the work actually needed, ideally with a cap.
What is a fixed capacity model? You pay an agreed fee for a team's time over a period, such as a month, and you choose what the team works on. It suits ongoing product work with a backlog. See the fixed capacity section above for what to write into the contract.
What does T&M mean in a quote, and what are T&M fees? T&M means time and materials. The fees are the hours worked, charged at agreed rates per role, plus materials or pass-through costs. Ask how time is recorded, how expenses and third-party costs are billed, and whether there is a cap.
Which is an example of a fixed-price contract? A marketing website with eight listed pages, agreed designs, and written acceptance checks, built for one agreed total and paid in instalments as each milestone is accepted.
What is the 40/20/40 rule in software engineering? It is an informal rule of thumb that splits effort into roughly 40% planning and design, 20% coding, and 40% testing. It is not a standard, but it is a useful reminder that a quote with little testing time is incomplete.
Is fixed price always safer for the buyer? It protects your budget for the written scope, but it does not protect you from a scope that turns out to be wrong. If you expect to change the product after user feedback, a fixed price can become a series of priced change requests.
Can I start with time and materials and switch to fixed price later? Yes. A common pattern is a short time and materials or fixed price discovery, followed by a fixed price or capped phase once the scope is specific enough to test.
Do I need a separate statement of work? Many agencies use a master services agreement for the legal terms and a statement of work for each project or phase. Check that the documents agree, and which one wins if they conflict.
What should I prepare before asking for a contract? A written brief with your users, tasks, integrations, must-pass checks, budget range, and deadline. The Diffwise.app project brief builder helps you write one, the free software project estimator gives an effort range and a suggested contract model for your scope, and you can contact Diffwise.app to discuss scope for a website, web app, or MVP.
Sources and further reading
- How to pay for Digital Outcomes and Specialists services (withdrawn)GOV.UK · Checked 18 November 2026
- The Digital, Data and Technology PlaybookGOV.UK · Checked 18 November 2026
- Manifesto for Agile Software Developmentagilemanifesto.org · Checked 4 November 2026
- Principles behind the Agile Manifestoagilemanifesto.org · Checked 4 November 2026
- How the discovery phase worksGOV.UK Service Manual · Checked 4 November 2026
- The Digital Services PlaybookU.S. DOGE Service · Checked 18 November 2026
- Circular 30: Works Made for HireU.S. Copyright Office · Checked 4 November 2026
- Copyright Law of the United States, Chapter 2: Copyright Ownership and TransferU.S. Copyright Office · Checked 4 November 2026
- Intellectual property and your work: what intellectual property isGOV.UK · Checked 4 November 2026
- How copyright protects your work: license and sell your copyrightGOV.UK · Checked 4 November 2026
- Copyright Act (R.S.C., 1985, c. C-42), section 13: ownership of copyrightJustice Laws Website, Government of Canada · Checked 4 November 2026
- A guide to copyrightCanadian Intellectual Property Office · Checked 4 November 2026
- Who owns intellectual property?IP Australia · Checked 4 November 2026
- Hiring contractorsbusiness.gov.au, Australian Government · Checked 4 November 2026
- FAR 16.207-1: Firm-fixed-price, level-of-effort term contractsU.S. General Services Administration, acquisition.gov · Checked 18 November 2026
- FAR 16.601: Time-and-materials contractsU.S. General Services Administration, acquisition.gov · Checked 18 November 2026
- FAR 16.202-2: Firm-fixed-price contracts, applicationU.S. General Services Administration, acquisition.gov · Checked 18 November 2026
