Fixed price vs time and materials: what's better for a web project?

Two stories that happen in this industry every week.

A mid-sized UK e-commerce business signs a time-and-materials agreement with a development partner. The estimate is £48,000 over fourteen weeks. By week sixteen, the invoice has crossed £92,000. The agency is still billing, the product is still not launched, and nobody is quite sure where the wheels came off. Each individual change felt reasonable. The aggregate didn't.

A SaaS founder, scarred by exactly that experience, signs a fixed-price contract for her next site. £62,000, twelve weeks, scope locked in writing. Halfway through user testing, she discovers the CTA on the pricing page is in the wrong place. The agency confirms they can move it. The change order is £1,400 and adds two weeks. The button is forty-eight pixels.

Both of these are real categories of failure, and they both come from picking a commercial model that didn't match the actual nature of the work. The question isn't which model is "better." The question is which model fits the project in front of you, and what to put in the contract regardless.

Fixed price, explained properly

A fixed-price engagement means the agency commits to a defined scope of work for a defined fee on a defined timeline. The price doesn't move unless the scope moves. Most reputable agencies build this around a written specification — a scope-of-work document that names every page, every component, every integration, every content type, and every acceptance criterion.

The thing people get wrong about fixed price is that it fixes scope, not outcomes. A fixed-price contract is the agency's promise to deliver the deliverables listed in the SOW. It is not a promise that those deliverables will produce the business result you wanted. That's a separate problem, and no commercial model solves it on its own. Outcome-based pricing exists, but it's vanishingly rare and almost always confined to performance marketing rather than build work.

Fixed price works best when the brief is genuinely well-defined. You know what pages the site needs. You know what the conversion flow looks like. You know which integrations are required. You have a strong design direction or you're prepared to defer to the agency's. The unknowns are small enough that the agency can price the work with reasonable confidence and you can sign off the spec with reasonable confidence that it covers what you need.

For the client, the risk is rigidity. The thing you didn't think of in week one — and there will be a thing — costs extra. Change requests run through a change-order process, and that process exists precisely because the commercial model demands it. The good agencies make change orders quick and frictionless. The bad ones make them adversarial, because every change is an opportunity to recover margin.

For the agency, the risk is the unknown unknowns. If a piece of work turns out to be three times harder than estimated, that's their problem. They priced it. They eat it. The way agencies manage that risk is by padding. A reasonable agency adds 15 to 25% contingency on top of the honest estimate. A less reasonable one adds 60% and pockets the difference when nothing goes wrong. You can spot the latter by getting a second quote and noticing that the line items are similar but the total is wildly higher.

Pricing models are risk-allocation models in disguise. Source: Pexels

Time and materials, explained properly

A T&M engagement means the agency bills you for the hours and resources they actually spend, against an agreed rate card. There's usually a not-to-exceed ceiling and a sprint cadence with regular check-ins. The scope can flex week to week. Whatever the team works on, you pay for.

T&M works best when the work is genuinely exploratory. Research-heavy projects, R&D, products where the requirements will change as you learn from users, complex integrations where you can't price the unknown until you've poked at it. In those situations, fixed price requires the agency to either over-estimate massively to cover the uncertainty (and you pay for risk you don't have) or to enter a project they'll lose money on (which never ends well for the client either).

The reason agencies prefer T&M is straightforward: it transfers commercial risk to the client. If the work takes twice as long, the agency still gets paid. They have no incentive to estimate accurately at the start, and limited incentive to be efficient mid-project. That isn't a moral failing of T&M agencies — it's the structural reality of the model, and any responsible agency working T&M actively counter-balances it with weekly burn reports, sprint demos, and pre-agreed budget guardrails.

The classic failure mode of T&M is the "open meter running" problem. There's no natural stopping point. Each week's invoice feels reasonable. The trajectory doesn't. By the time you realise you've spent £80,000 and the product is 60% done, the sunk-cost decision is unpleasant: keep paying or kill it and lose what you've spent. The way to avoid that is to set hard budget gates upfront — "we review at £30K, we review at £60K, we don't cross £75K without a written extension."

Hybrid pricing trades a single hand-off for many small ones. Source: Pexels

A third option: milestone-based hybrid

The honest answer for most complex projects is a hybrid. You fix the phases, you T&M within each phase, and you put decision gates between them.

Discovery is T&M against a small fixed cap — usually two to three weeks. The output is a real specification document. Then design is fixed-price against that spec. Then build is fixed-price against the signed-off design. Inside each phase, the team flexes how they spend their hours; between phases, the client signs off and re-confirms scope before money commits to the next stage.

This works because it matches the model to where the uncertainty lives. The earliest phases are the most uncertain, so they're the most flexible. By the time you're building production code, the uncertainty has been priced out, and a fixed quote is sensible.

It's also harder to sell on the first call, because it requires the client to commit to a process rather than to a number. But for projects above about £40,000 with any meaningful complexity, it produces better outcomes than either of the pure models.

How to choose

Pick fixed price when: the brief is well-defined, the outcomes are clear, the budget is fixed, the integrations are known, and the timeline is firm. If you can hand the SOW to two agencies and get comparable quotes, fixed price is the right model.

Pick T&M when: you're doing research-heavy or exploratory work, the requirements will evolve, the team needs to adjust to what they learn, and you have the internal capacity to actively manage the burn rate.

Pick a hybrid when: the project is large, the discovery phase will materially change the build, and you want commercial discipline without locking yourself into a spec you wrote before you knew enough to write it well.

The right pricing model is the one both parties can defend on a bad day. Source: Pexels

What protects you regardless of model

A clear scope document. Not the proposal — the scope document. Most disputes between agencies and clients come down to one side reading the proposal as a list of marketing promises and the other side reading it as a contractual scope. Get a single document signed by both sides that names every deliverable and every acceptance criterion. If a page is mentioned in the proposal but not in the SOW, it's not in scope.

A defined change-order process. Agreed upfront. "Any change of more than X hours triggers a written change order with a fixed fee and a revised timeline." That sentence prevents 90% of fixed-price disputes. It also prevents the T&M version of the same dispute, which is "wait, when did we agree to that?"

Weekly demos. Not status updates. Demos. The agency walks the client through what was actually built that week. The client can see drift in real time. The agency can flag concerns before they compound. Both models break when the first time the client sees the work is at delivery.

Decision gates. A small number of formal sign-off points where the client confirms a phase is done and commits to the next one. These are particularly important on hybrid engagements but useful on any of the three models.

🎬 Watch: Scoping a project together

A pricing model is something you negotiate openly — or pay for later.

How Nexus structures fixed-scope engagements

For most of our work, we use fixed scope with built-in flex. Here's what that means in practice.

We start with a paid discovery phase — usually two weeks, sometimes three. The output is a written SOW, an architectural plan, a content map, and a real estimate. The discovery fee is fixed. Either side can walk away at the end of discovery without obligation to proceed.

The main build is then fixed-price against that SOW. We carry our own contingency in the number — we don't surface it as a separate line item. The number you sign is the number you pay, unless you ask for something that wasn't in the spec.

Change orders are explicit and quick. If you want something added, we tell you the cost and the timeline impact in writing within 48 hours. You decide. We don't sit on change orders as leverage.

We hold a weekly demo with whoever's running the project on your side. No status decks. We show you what we built, you tell us what's working and what's not, we adjust. By the time we hit launch week, there are no surprises.

And we offer a small T&M envelope post-launch for the work that always exists after a build but is impossible to price upfront — the refinements, the small additions, the optimisations as you learn from real traffic. That's separate from any ongoing maintenance plan, and it stops when you say it stops.

Get a clear quote

If you've got a brief that's well-defined enough to price, send it to us. We'll respond with a fixed-scope quote that names what's in, what's out, and what it costs. If the brief isn't quite there yet, we'll propose a short discovery instead. Either way, you'll know what you're paying for before you sign anything.

Start the conversation.