Most conversations about commercial models go wrong in the first thirty seconds. Someone from procurement or finance frames it as a question about cost control. "We need to know what this is going to cost us." And from that moment, the entire discussion is about price rather than about where the risk sits, who carries it, and whether the agency you're about to pay has any financial reason to care about the same things you care about.
I sat in a steering committee meeting about eighteen months ago - a professional services firm, around 300 people, about to commission a significant piece of digital work. The COO opened with: "We'd like this done on time and materials so we have full transparency." And the agency nodded enthusiastically. Of course they did. Because T&M is the model where the agency gets paid for showing up, regardless of what gets delivered. The project ran for nine months. The original estimate was £180k. Final invoice: just over £290k. When the COO asked what had changed, the agency produced a spreadsheet of hours. Every one of them legitimate. Every one of them real. Not a single one of them explained why the outcome was worse than what a fixed-price engagement would have produced for £40k less.
Transparency of hours is not the same as transparency of value. But it sounds like it is, and that's the problem.
We've always done time and materials. It's transparent and we know what we're paying for.
I understand the appeal. Genuinely. You can see every hour billed, query any line item, and it feels like you're in control. But you're confusing visibility with governance. You can see what you're paying - you just can't see whether what you're paying for is the most efficient path to the outcome you actually need. I've written about why hourly billing rewards the wrong behaviours separately, and that piece makes the fuller case for why the incentives are structurally misaligned. This piece is the companion to that one - a practical decision framework for someone who's about to structure a commercial engagement and wants to get it right.
So let's walk through the options properly. Not as an academic exercise, but as a working tool.
The case for fixed price is genuinely strong in the right circumstances. You agree a scope, you agree a price, the agency delivers against both. Your budget is protected. The agency is incentivised to be efficient because every hour they save is margin they keep. Clean, predictable, accountable.
The downsides are well understood. If requirements change - and they almost always change - you're into change request territory, which can get adversarial quickly. For discovery-heavy work, or anything where you genuinely don't know what you'll find once you start digging, fixed price can feel like buying a house based on the estate agent's photos.
But there's a risk that doesn't get discussed enough, and it's the one I want you to sit with.
When an agency is delivering a fixed-price project and the margin starts getting squeezed - maybe the scope was slightly underestimated, maybe the third round of amends took longer than planned, maybe the client's internal review process added two weeks - the agency faces a choice. Absorb the cost and deliver to the original quality standard. Or make small, invisible compromises that protect the margin. Corners get cut. Testing gets compressed. The content migration that was supposed to be reviewed manually gets run through an automated script with a quick spot-check. The accessibility audit becomes a checklist rather than a proper evaluation.
These aren't dramatic failures. They're the quiet erosion of quality that you won't discover until after handover - maybe not for months. A firm comes to us after a fixed-price project with another agency, and when we start looking under the bonnet, we find technical debt that was baked in during the final weeks when the agency was trying to close the project without blowing their margin. It's more common than I'd like it to be.
Fixed price works brilliantly when the scope is well-defined, the requirements are stable, and you have a mature relationship with an agency you trust to flag problems early rather than hide them. Choose it purely for cost certainty and you're picking the right model for the wrong reasons.
T&M gives you something genuinely valuable: the ability to change direction without a formal change control process. If you're in the early stages of a project where the requirements are uncertain - a discovery phase, a prototype, an exploration of what's technically feasible - that flexibility matters. You don't want to be locked into a scope that was defined before you understood the problem properly.
The transparency argument is real too, up to a point. You can see every hour, every task, every team member's time. It feels like control.
But the structural problem with T&M is that the agency's revenue is directly tied to the number of hours worked. Not to the efficiency of those hours. Not to the quality of the output. Not to the business outcome you're trying to achieve. Hours worked. That's the incentive. And incentives, over time, always win.
Surely a good agency just does the right thing regardless of the model?
Some do. Most try to. But good intentions and structural incentives are different things. Without a price ceiling, there's no natural forcing function for efficiency. Problems that could be solved in a day get an extra half-day of polish. Meetings that could be thirty minutes run to an hour. Nobody's acting in bad faith. The commercial model just doesn't reward speed.
The governance burden also falls entirely on you. You need someone internally who can evaluate whether the hours being billed represent productive, efficient work or whether the project is drifting. Most mid-sized firms don't have that person - or if they do, they have seventeen other things competing for their attention.
I've seen T&M projects where the total spend ended up 40-60% higher than what a fixed-price engagement would have cost for the same scope. Not because anyone was dishonest, but because the model had no mechanism to say "that's enough." The COO from that steering committee? That was one of those projects.
T&M is appropriate when the scope is genuinely uncertain, when you have strong internal governance capability, and when the relationship with the agency is established enough that you trust their judgment on how to spend your money. For a first engagement with a new agency? It's a brave choice. Braver than most people realise.
On paper, outcome-based pricing is the dream. The agency gets paid when you get results. Perfect incentive alignment. Why would anyone choose anything else?
In practice, it's genuinely difficult to make work for most digital projects. And I say that as someone who believes deeply in incentive alignment.
The first problem is definition. What, exactly, is the outcome? A 30% increase in qualified enquiries? Sounds clear enough - until you need to define "qualified," agree the attribution model, account for the SEO work another agency is doing simultaneously, and decide what happens if the outcome is partially achieved. We hit 22% instead of 30% - is that a success or a failure? What does the agency get paid?
The second problem is attribution. Most digital outcomes are influenced by multiple factors. The agency redesigned your website and rebuilt your conversion paths, sure. But your sales team also started following up on enquiries faster, your competitor had a PR crisis, and you published that thought leadership piece that went viral on LinkedIn. How much of the improvement is the agency's work? This isn't a theoretical question - it's the question that ends up in a tense meeting six months into the engagement.
And then there's the problem that bothers me most. When an agency is paid for a specific measured outcome, they will optimise for that outcome. Rationally, logically, predictably. If the outcome is enquiry volume, they'll drive quantity at the expense of quality. If it's conversion rate, they might tighten the funnel in ways that reduce total reach. You get what you measure, and you get only what you measure. The unmeasured dimensions - brand perception, content quality, code maintainability, accessibility - become secondary because they're not what the agency is being paid for.
Outcome-based works when the outcome is narrow, well-defined, genuinely attributable to the agency's work, and measurable within a reasonable timeframe. It works best in mature relationships where there's enough trust to handle the grey areas without lawyers. For most digital engagements, it's not the right primary model. But it can be a brilliant component of a hybrid structure - which is where this actually gets interesting.
I know. "The answer is a hybrid" sounds like the consulting equivalent of "it depends." Bear with me, because the specific structures matter.
Phased fixed-price is the model I recommend most often, and it's worth showing what it actually looks like rather than just naming it. Take a £200k website rebuild. Rather than fixing the entire scope upfront, you structure it as three phases: a 14-day assessment at around £15k, a first delivery sprint covering architecture, core pages, and conversion paths at around £90k, and a second sprint covering content migration, integrations, and optimisation at around £95k. At the end of each phase, there's a gate review - a half-day session where both sides assess what's been learned, what's changed, and what the next phase should contain. The COO signs off before anything moves forward.
What this gives you: cost predictability within each phase, the ability to course-correct between phases, and a natural moment to reassess whether the original brief still makes sense. We use this structure at Distinction - it's how our WHNN framework operates in practice, with quarterly cycles that let you adjust course without losing momentum. The gate review is the bit most firms skip when they try to run this themselves, and it's the bit that makes the whole thing work.
T&M with a cap gives you flexibility within a boundary. The agency bills for actual hours worked, but there's a ceiling. If the project can be delivered for less, you pay less. If it hits the cap, the agency has to make a decision about how to deliver the remaining scope within the budget. It's not perfect - you still carry some of the governance burden of T&M - but the cap introduces the forcing function that pure T&M lacks.
Retainer with objectives is the most underused model I know. Instead of buying a block of hours each month, you agree a set of outcomes or deliverables for a monthly fee. The agency is incentivised to deliver those outcomes efficiently because the fee is fixed regardless of hours worked. It works best for ongoing relationships where the agency's contribution is continuous - managing and optimising a platform, running a content programme, providing ongoing strategic advisory. We've structured several of our longer-term client relationships this way, and the quality of the work is noticeably different from pure hourly arrangements. The agency thinks about what to do, not how many hours to bill.
Three questions. Simple, but they work.
How well-defined is the scope? If you can write a detailed brief that you're confident won't change significantly, fixed price is your friend. If you're still figuring out what you need, T&M or phased fixed-price. If you're somewhere in between - which is where most projects sit - phased fixed-price gives you the best of both.
How measurable and attributable is the outcome? If you can point to a specific, measurable result that is clearly driven by the agency's work, outcome-based pricing (or a hybrid with an outcome-based component) deserves serious consideration. If the outcome is complex, multi-factor, or influenced by things outside the agency's control, avoid it. The attribution arguments aren't worth the damage they do to the relationship.
How mature is the relationship? This is the one that gets forgotten. If this is your first engagement with a new agency, start with fixed price for the first phase. It establishes trust, demonstrates capability, and gives both sides a shared reference point before you commit to anything more open-ended. Any agency pushing hard for T&M on a first engagement is a yellow flag - not necessarily a red one, but worth interrogating. Once you've worked together and built trust, the full range of models becomes available.
If you want a quick reference for this, we've put together a commercial model decision guide - a one-page tool that maps these three questions to the appropriate model, including the hybrid options. It's designed for whoever in your firm is structuring the next agency engagement. You can download it [here].
The commercial model you choose is a governance decision. It determines where the risk sits, where the incentives point, and whether you and your agency are actually playing the same game.
Default to T&M and you're consistently choosing the model that rewards utilisation over outcomes. Default to fixed price without proper scope definition and you're creating the conditions for invisible quality compromises. Chase outcome-based pricing without the measurement infrastructure and you're building in a dispute you'll have six months from now.
Match the model to the project, the relationship, and the risk profile. And if you're not sure which model fits a specific engagement you're about to structure, book a commercial model advisory conversation with us - thirty minutes and we'll give you a direct recommendation. No pitch, just an honest answer.
Getting this right isn't about saving money on the engagement. It's about making sure the money you spend actually buys what you think it's buying.