←
BACK_TO_BLOG

Fixed Price vs Time and Materials vs Sprint-Based Pricing

Share
X_/_TwitterLinkedInCopy_Link
TL;DR
  • A fixed price for the whole project only works when the scope can be written down completely before work starts. For a first build, it usually can't, so the vendor pads the quote or fights you on every change.
  • Time and materials bills the hours actually worked, which suits work nobody can estimate yet. It needs someone on your side who reads the timesheets every week, and an owner running the business rarely has those hours.
  • A fixed price per two-week sprint caps each payment and lets the scope move between sprints. You see working software every two weeks and decide whether to buy the next sprint.
  • Ask any vendor what happens at the end of a sprint you didn't like. If the answer is a cancellation fee or a long notice period, the sprint price is a fixed-price project in disguise.
A row of cash envelopes standing like dominoes, each with a blue paper screen on its front. Open scissors have cut the first envelope off from the rest, and a bright green strip lies across it.

The fixed price vs time and materials choice is a choice about who carries the risk of not knowing. A fixed price puts it on the vendor, who charges you for carrying it. Time and materials leaves it with you, billed by the hour. For a small business commissioning its first custom software, our answer is a third model: a fixed price for each two-week sprint, with the right to stop after any one of them.

Below is the best case for each side, taken from the US federal contracting rules and the government's own guide to buying software, then where sprint pricing goes wrong and when we'd tell you to pick something else.

The Case for a Fixed Price Software Contract

A fixed price contract for software development sets one price for an agreed scope, and the vendor absorbs any overrun. The US Federal Acquisition Regulation describes the firm-fixed-price contract as a price "not subject to any adjustment on the basis of the contractor's cost experience". It "places upon the contractor maximum risk and full responsibility for all costs and resulting profit or loss."

For an owner, that is the appeal. You know the number before you sign, and a vendor who underestimated eats the difference.

The same regulation is clear about when that works. A fixed price is suitable for work bought "on the basis of reasonably definite functional or detailed specifications", when a fair price can be set at the outset.

Some small-business projects meet that bar. Moving a working quote calculator from a spreadsheet to a web page, with the same inputs, the same rules and the same PDF out, is reasonably definite. So is a second version of software you already run, where every screen exists and the change list is written down.

A first operations system doesn't meet it. The rules live in people's heads, and the first time staff use the screens, they find steps nobody mentioned. Every one of those becomes a change request on a fixed-price contract, priced by a vendor who now knows you can't walk away.

The Case for Time and Materials

A time and materials contract pays a team's hourly rates for the hours they actually work, plus any materials. It is the model the US government's 18F team recommended for custom software, in a guide archived in January 2025. Its guide to buying software development services pairs a statement of objectives with a time and materials contract and a "not-to-exceed" ceiling, used to "buy the time and expertise" of a team.

18F's argument is that for software projects this "protects the government's financial and performance interests better than other contract types." The vendor bills "only for actual time incurred", every invoice carries exact billing data, and fixed-price invoices "lack detail" by comparison.

The regulation is narrower about when to use it. Time and materials may be used "only when it is not possible at the time of placing the contract to estimate accurately the extent or duration of the work". That describes a first build well.

The same paragraph names the cost. A time and materials contract "provides no positive profit incentive to the contractor for cost control or labor efficiency", so the buyer has to watch the work. 18F's answer is a product owner employed by the buyer, who "must be available to the team and empowered by the organization" and is assigned at least half-time to the project.

A half-time product owner is the condition a small business usually can't meet. The owner signing the contract is also running dispatch, chasing invoices and covering for whoever called in sick. Nobody on the payroll can tell whether 30 hours on "user permissions" was fair.

Why Fixed Price per Sprint Suits a First Build

A fixed price per sprint takes the useful half of each model. The scope stays open, as in time and materials, and each two-week block has a set price, as in a fixed-price contract.

Here is how it runs in practice. Before each sprint, you and the team agree what it will deliver. At the end, you see that work running, on a screen, with your data. Then you choose whether to buy the next one.

The risk you carry is capped at one sprint's price. On our pricing page that cap is $2,500, and the team shows its work at a demo every week of the sprint. You still have to turn up to the demo, but you're judging working screens, and nobody has to audit a timesheet. When we priced a three-month first version for a prospective client this July, we wrote it as six two-week sprints, invoiced one sprint at a time.

The 2001 Agile Manifesto put the idea in one line: "Customer collaboration over contract negotiation". A sprint price makes that line cheaper to keep, because nobody has to renegotiate a contract to change next fortnight's plan.

An old balance scale with a padlocked bundle of papers on the lower left pan, a heap of stopwatches on the right pan, and a blue paper square with a bright green strip hanging from the centre post

How much of your build is written down?

  • SCOPE??
  • SPRINT GOAL??
  • DEMO??
  • DECIDE??
  • NEXT SPRINT??

Tell us which rules your team keeps in its head. In one call, we'll mark what is definite enough to build in sprint one and what should wait.

SCOPE_MY_FIRST_SPRINT→

Where Sprint-Based Pricing Goes Wrong

Sprint pricing caps each payment. The total can still grow without a limit, one reasonable-looking sprint at a time. Three habits keep the total in check.

Ask for a sprint count up front. A vendor who can't estimate the number of sprints, even roughly, doesn't understand the work yet. Our post on how much custom software costs shows how to turn a sprint count into a total and why to hold back a reserve on top.

Write down what "done" means for each sprint. A sprint that ends with "progress on the back end" has nothing for you to check. Ask for something you can click.

Read the exit terms. If stopping after a bad sprint costs a notice period or a fee, you've bought a fixed-price project paid in instalments. The sprint price only protects you if leaving is cheap, and if the code you've paid for is already yours when you leave.

Scope still needs writing down. Sprints let it change, and they don't replace the work of deciding it. A one-page PRD before sprint one saves at least one sprint of discovering rules the hard way.

When to Choose a Fixed Price or Time and Materials Instead

Honestly, sprints aren't always the answer. The verdict flips in three cases.

The scope is genuinely definite. If you're rebuilding a tool that already works, with every rule written down, a single fixed price is fair to both sides. Ask the vendor to list what's excluded, because that list is where the change requests will come from.

You have a technical person in-house. An operations manager who used to be a developer, or a part-time CTO, can be the product owner 18F asks for and can also judge whether the hours billed were fair. Then time and materials with a not-to-exceed ceiling buys you the most flexibility for the money.

The software is live and the work is small and steady. Bug fixes and small changes after launch don't fit two-week goals. A monthly retainer or hourly support with a cap suits them better.

If none of those three cases is yours, what would you need to see at the end of sprint one to buy sprint two?

FAQs

// QUESTIONS //
Q.

What is a not-to-exceed clause in a software contract?

A not-to-exceed clause sets the most the client can be billed under an hourly contract. The US federal rules require a ceiling on every time and materials contract, which the contractor exceeds at its own risk.

Q.

How long is a sprint in software development?

The Scrum Guide defines sprints as "fixed length events of one month or less". We use two weeks, so an owner gets a demo and a decision twice a month.

Q.

Can a fixed price project use agile methods?

It can if the contract allows scope swaps within the fixed price, trading one feature out when a new one comes in. Without that clause, every change becomes a priced change request.

Q.

Who owns the code on a sprint-based contract?

Ownership depends on the assignment clause in the contract, whatever the pricing model. Our post on who owns AI generated code covers what that clause should say.

2-WEEK SPRINTS
//
$2,500 FLAT
//
LIVE IN 90 DAYS
//
THE CODE IS YOURS
//
ZERO LOCK-IN
//
2-WEEK SPRINTS
//
$2,500 FLAT
//
LIVE IN 90 DAYS
//
THE CODE IS YOURS
//
ZERO LOCK-IN
//
2-WEEK SPRINTS
//
$2,500 FLAT
//
LIVE IN 90 DAYS
//
THE CODE IS YOURS
//
ZERO LOCK-IN
//
2-WEEK SPRINTS
//
$2,500 FLAT
//
LIVE IN 90 DAYS
//
THE CODE IS YOURS
//
ZERO LOCK-IN
//

Pay for two weeks. [See what you got.]

START_PROJECT
→