Fixed-price vs time-and-materials for custom software

Quick answer: fixed-price suits tightly-scoped, well-understood projects with clear acceptance criteria. Time-and-materials suits genuinely uncertain scopes where the right answer emerges through the work. Fixed-price on undefined scope produces change-request hell; T&M on well-understood scope produces aimless drift. The hybrid that usually works: fixed-price for a tightly-scoped phase one, T&M (or a fresh fixed-price) for whatever phase two looks like.

The contract-model question comes up early in every engagement. Both models are legitimate. Both have failure modes. The choice between them is more important than most clients realise — not because one is universally better, but because mismatching the model to the project shape is the source of most painful agency relationships.

Where fixed-price wins

Fixed-price is the right model when:

1. The scope is tightly defined and well-understood by both sides. A documented brief with explicit acceptance criteria. Functional requirements that aren’t expected to change during the build. The kind of project where you could write down what “done” looks like in advance and both parties would agree.

2. The work pattern is familiar. A marketing site of the kind we’ve built before on a stack we know well. An ecommerce build on a familiar platform. A migration of a known type. The estimating risk is bounded because we’ve seen the shape before.

3. The client wants budget certainty. Some clients have hard budget ceilings driven by financial planning, board approvals, or grant funding. For these clients, fixed-price gives the certainty they need to commit. The certainty has a cost (the agency builds in margin for risk), but for the right project, that’s a fair trade.

4. The agency has the experience to estimate accurately. A team that’s built fifty similar projects estimates more accurately than one building its first. For agencies in the “built fifty” category, fixed-price on familiar work is reasonable for both sides.

The signal: if the team can produce a confident, structured estimate that breaks down phase-by-phase and feature-by-feature, with explicit risk allowances, the project is probably suited to fixed-price.

Where fixed-price breaks

Fixed-price is the wrong model when:

1. Scope is genuinely undefined. “Build us a custom system to manage our operations” is not a fixed-price brief. The team can’t estimate what hasn’t been specified. If the agency quotes a fixed price anyway, one of two things happens: either the price is wildly inflated to cover the unknowns, or the change requests during the build absorb the actual cost.

2. Discovery hasn’t happened. Fixed-price after a serious discovery phase can be reasonable. Fixed-price before discovery, with the discovery folded into the build, is usually fiction. The agency is quoting what they hope it will cost; the client is committing to what they hope they’re getting.

3. The technology or domain is genuinely new to the team. A team that’s never built an AI product, never integrated with a specific complex platform, never worked in a regulated industry — their estimate is a guess. Fixed-price on unfamiliar work is how agencies lose money, or how clients pay for the agency’s learning curve.

4. The client has many stakeholders with diverging views. Fixed-price assumes a stable scope. With many stakeholders who haven’t aligned, the scope keeps shifting throughout the build. Each shift is a change request, each change request is a negotiation, and the relationship deteriorates regardless of who’s being “reasonable.”

The failure pattern is consistent: a fixed-price contract on undefined or unstable scope produces months of arguing about what was “in scope” and a final product that resembles the brief but feels brittle. Both parties end up unhappy.

Where time-and-materials wins

T&M is the right model when:

1. The scope is genuinely uncertain and the team needs to discover it through the work. Some projects fundamentally can’t be scoped fully up-front. Research-heavy work. Complex integrations where the requirements emerge from the data. Projects where the right architecture only becomes clear once a v1 is running. T&M lets the work proceed without forcing the agency to over-promise.

2. The client has the maturity to operate T&M well. This includes: weekly visibility into hours and progress, the ability to make scope decisions iteratively, an understanding that the budget is a guideline that can flex, and the discipline to stop work when the value runs out. Without these, T&M becomes uncapped spending.

3. The project is genuinely high-trust. T&M works best between parties who’ve worked together before, or where references and reputation make the trust durable. New relationships often benefit from fixed-price as a forcing function for clear communication, even when T&M would be theoretically better.

4. The right answer changes as the work progresses. For genuinely exploratory work — new product development, prototyping, R&D — the original brief is often wrong by week three. T&M lets the work pivot without requiring contract amendments.

For more on the structural risks of fixed-price contracts on undefined scope, the PMI’s perspective on contract types is a sound reference covering the same patterns from a project-management lens.

Where T&M breaks

T&M is the wrong model when:

1. The client doesn’t have the operational discipline to manage it. T&M without weekly visibility, scope decisions, and budget discipline drifts. The agency keeps working, the bill keeps growing, the client doesn’t know whether the work is on track. Eventually a confrontation. Always avoidable in principle, often unavoidable in practice.

2. The project is well-understood and the work is routine. For a known shape of work, T&M shifts risk that the agency can comfortably absorb onto a client who can’t. The right model for routine work is fixed-price — the agency carries the estimating risk because they’re positioned to.

3. The client needs budget certainty. T&M can’t guarantee a number. For clients who genuinely need a number, T&M is structurally wrong — even with caps, even with weekly burn rates, the model doesn’t deliver what they need.

4. The relationship is new and trust is low. A first project together is often a poor T&M candidate. Both parties are still learning how to work with each other; the structure of fixed-price helps. Subsequent projects between the same parties are different.

The hybrid that often wins

For complex projects, the model that consistently produces good outcomes:

Phase 1: tightly-scoped fixed-price. After discovery, define a phase one with deliberately narrow scope. Fixed-price it — the team can estimate this confidently because the scope is defined. Ship it within 6–10 weeks. Both parties learn from the work.

Phase 2 onwards: re-decide based on what you’ve learned. After phase one, the project landscape is clearer. Maybe phase two is also tightly-scoped and worth a fresh fixed-price. Maybe phase two is genuinely uncertain (the data turned out different than expected) and T&M suits. Maybe the project is done and phase two doesn’t happen.

This pattern lets fixed-price’s certainty work for the parts of the project that are knowable, while letting T&M’s flexibility work for the parts that aren’t. It also creates a natural decision point after phase one: continue, change direction, or stop.

Reading a fixed-price quote honestly

A fixed-price quote that’s genuinely reasonable for the work usually has:

• A breakdown of phases / features with hours or days against each
• An explicit list of what’s in and out of scope
• A documented assumption set (what the agency assumed when estimating)
• A clear change-request process for things outside the scope
• A risk allowance that’s honestly named, not buried

A fixed-price quote that’s a wager usually has:

• A single bottom-line number with no breakdown
• Vague scope language (“modern, professional design”, “essential features”)
• No documented assumptions
• Aggressive timelines without buffer
• Suspiciously low pricing (which is recovered through change requests later)

Reading these signals is most of the value of comparing quotes. The cheapest number isn’t always the cheapest project.

What we do

For client work at our studio:

• We default to fixed-price for projects after discovery has produced clear scope. Most marketing sites, most known integrations, most ecommerce builds.
• We default to T&M for genuinely uncertain work, especially research-heavy or AI-heavy projects where requirements emerge.
• We use the hybrid (phase-one fixed, later phases re-decided) for larger custom builds where total scope is too large or too uncertain to fix up-front.
• We refuse fixed-price on projects where discovery hasn’t happened yet — we’ll do discovery first as its own engagement, then quote phase one based on what discovery surfaces.

This isn’t universal. Other reasonable agencies operate differently. But the pattern is consistent: matching the model to the project shape, rather than insisting on one model for every engagement.

Common questions

Is fixed-price or time-and-materials better? Neither is universally better. Fixed-price suits tightly-scoped, well-understood projects. T&M suits genuinely uncertain projects with mature clients. Mismatching the model to the project produces predictable failures — either change-request hell (fixed-price on undefined scope) or aimless drift (T&M on routine work).

Why do agencies prefer time-and-materials? Some do (lower estimating risk, more flexibility), some don’t (worse cash flow predictability, harder to sell to budget-constrained clients). The honest agencies use both, matching to the project. The agencies that only do one are usually optimising for their own operations rather than the client’s situation.

Can I get a fixed price for an undefined project? You can usually get a number. Whether it’s a real fixed price is the question. A fixed price on undefined scope is structurally either inflated (to absorb the unknowns) or fictional (with the actual cost emerging through change requests). The honest path is to do discovery first, then fix-price the defined work.

What’s a typical contract structure for a large custom build? Often a hybrid: discovery as a fixed-price engagement (1–4 weeks), then phase one as a fixed-price build (6–12 weeks), then later phases re-decided based on what’s learned. This pattern combines the certainty of fixed-price with the flexibility T&M offers, applied where each fits.

Should I be worried about a very low fixed-price quote? Yes, usually. A quote materially lower than competitors often reflects either underscoping (the agency hasn’t engaged with the brief) or change-request strategy (the build cost is recovered through scope additions). Read the breakdown, the scope language, the assumption set, and the change-request process — the cheap number is rarely the cheap project.

If you’re structuring a contract for a custom build and unsure which model fits, start a project and we’ll walk through it honestly — including refusing the engagement if the model is wrong for the project shape.