Fixed price vs time and materials for software testing

Updated 2026-09-24

Fixed price suits testing work you can specify in advance, such as a defined regression pack or an accessibility audit against a named standard. Time and materials suits ongoing testing where scope moves with your roadmap. Under a fixed price, the vendor absorbs overruns by testing less, so define depth and acceptance criteria. Under time and materials, you carry the risk, so add a ceiling, monthly reporting against it and measurable service levels.

What to check

  1. Decide whether you can write down the scope, depth and acceptance criteria before the work starts; if not, fixed price will be priced for risk.
  2. For fixed price, define test depth in writing: which features, which platforms, which types of testing, and what counts as done.
  3. For fixed price, check the change control clause, since every scope change becomes a priced change request.
  4. For time and materials, set a ceiling per month or per phase that the vendor cannot exceed without written approval.
  5. For time and materials, require timesheets by person and activity and a monthly report against the ceiling.
  6. Under either model, attach measurable service levels so you are paying for outcomes as well as effort.

Ask the vendor

  • Is your price based on time and materials, fixed price or outcomes, and exactly which activities are inside the price and which are billed separately?
  • If we pay for time, how are hours recorded, approved and evidenced on invoices, and can we see timesheets by person and activity?
  • How do you handle change requests: who can raise them, how are they estimated and approved, and what happens to work that starts before approval?
  • Under a fixed price, what do you do if the testing takes longer than planned?
  • Would you accept tying part of the payment to acceptance of deliverables or milestones?

Two ways of sharing risk

The choice between fixed price and time and materials is a choice about who carries the risk that the work takes longer than expected. Under a fixed price, the vendor does. Under time and materials, you do. Neither is right for all testing work. The question is which risk you can manage.

US federal acquisition rules describe the trade-off clearly. A firm fixed price is suitable when the work has reasonably definite functional or detailed specifications and the contractor can identify the uncertainties and accept the risk. A time and materials contract may be used only when it is not possible to estimate the extent or duration of the work accurately, it must include a ceiling price, and because it gives the contractor no positive profit incentive for cost control or labour efficiency, it requires appropriate surveillance of performance by the buyer. Those three points, specification, ceiling and oversight, are the practical guide for a QA contract too.

When each model fits testing work

Testing work Usually fits Why
A defined regression pack for a stable release Fixed price Scope and depth can be written down
An accessibility audit against WCAG 2.2 at a named level Fixed price The standard defines the scope
An initial automation suite with acceptance criteria Fixed price or capped time and materials Can be defined if the acceptance criteria are clear
Ongoing testing of each sprint's changes Time and materials or a monthly fee Scope follows your roadmap
Exploratory testing of a new product Time and materials You learn the scope as you test
A performance test for a launch Fixed price if workload models exist, otherwise capped time and materials Depends on how much is known upfront

How a fixed price squeezes test depth

When a fixed-price testing job runs long, the vendor's options are to lose money or to test less. Testing less is hard for you to see. The test cases still run, but with fewer data variations, fewer platforms, shallower exploratory sessions or less time spent reproducing an intermittent defect.

Protect yourself by defining depth, not only scope:

  • the features and user journeys in scope;
  • the browsers, devices and operating systems;
  • the types of testing, such as functional, exploratory, accessibility or performance;
  • the test design techniques or coverage expected for high-risk areas;
  • the entry and exit criteria for the cycle;
  • the acceptance criteria for each deliverable.

Then read the change control clause. Under a fixed price, anything outside the written scope becomes a change request. The framework's change control clause requires the vendor to state the effect on scope, charges, timetable and service levels before any change takes effect, and prevents the vendor from quietly shrinking the team to absorb a change.

How to cap time and materials

Time and materials gives you flexibility. You can move testers to the feature that matters this sprint without a contract change. The cost is that the vendor has no built-in reason to be efficient. Add the controls yourself:

  • A ceiling. A monthly or phase limit on hours or spend that needs your written approval to exceed.
  • Timesheets with detail. Hours by person and by activity, not a single monthly total.
  • Monthly reporting against the ceiling. Spend to date, forecast to the end of the period, and the reason for any variance.
  • Rate lock. Fixed rates for the initial term, with a capped review after that.
  • Seniority control. No change to the mix of roles that affects charges without your approval.
  • Service levels. Measurable targets, such as regression completed within an agreed time of a release candidate, so you pay for results as well as hours.

The framework's clause on service levels asks that each one state what is measured, the target, the data source, the measurement period and who reports it. Performance-based acquisition rules in US federal procurement follow the same idea: describe the work in terms of required results, with measurable performance standards.

Comparing quotes across models

Vendors will often quote in different models, which makes comparison hard. Convert each quote into a monthly and an annual figure for the same scope, including the change requests you expect under a fixed price and the likely hours under time and materials. The cost model supports time and materials, monthly fee per person, fixed price per release and base plus variable fee, and compares them on your numbers.

Then write the pricing terms into the contract. The contract clause builder produces the rate lock, change control, service level and acceptance clauses. For a wider view of pricing structures see QA pricing models compared and rate cards and what they hide.

Tools for this

Questions buyers ask

Is fixed price cheaper for testing?
Not necessarily. A vendor pricing fixed-price work adds a margin for the risk it carries. If the scope is uncertain, that margin can exceed what time and materials would have cost. If the scope is clear and stable, fixed price can be predictable and fair.
Can we combine the two models?
Yes. A common pattern is a fixed price for a defined piece of work, such as building an initial automation suite with acceptance criteria, followed by time and materials or a monthly fee for maintenance and ongoing testing.
What about outcome-based pricing?
Some vendors offer a base fee plus a variable part tied to measured outcomes. It can align incentives, but only if the outcomes are ones you can verify yourself. See the KPI guide before agreeing to one.

Sources

  1. FAR Subpart 37.6: Performance-Based Acquisition, US Federal Acquisition Regulation
  2. The Sourcing and Consultancy Playbooks, UK Cabinet Office
  3. TechFAR Hub, US Digital Service
  4. Model Services Contract, UK Cabinet Office
  5. FAR 16.601 Time-and-materials contracts, US Federal Acquisition Regulation
  6. FAR 16.202-2 Firm-fixed-price contracts: application, US Federal Acquisition Regulation