Every agency has an opinion about pricing that happens to suit the way it likes to work. Here is the version with the incentives made visible, including ours.
What fixed price actually means
You agree a scope, a number, and a date. The agency carries the risk of getting the estimate wrong.
That risk is real, so it gets priced in. A fixed-price quote almost always contains a buffer, often 20 to 40 percent, because the agency is insuring itself against its own estimation error. You are paying a premium for certainty, the same way you would for any insurance.
It works well when the scope is genuinely knowable up front: a rebuild of something that already exists, a well-understood integration, a design system, a marketing site.
It goes wrong when the scope was never knowable. Then the buffer runs out mid-project and you get the thing every client dreads: an agency arguing that your reasonable request is "out of scope," because at that point every extra hour comes out of its margin. The incentive has quietly flipped from build the right thing to build exactly the listed thing, quickly.
What time and materials actually means
You pay for the hours worked. You carry the risk of the estimate being wrong.
It works well when discovery is genuinely open-ended, when you have an internal product owner who can steer week to week, and when you trust the team enough not to need a contractual cage.
It goes wrong when nobody is watching the burn. Without a cap and a cadence, T&M rewards slow work, not through malice, usually, but because there's no forcing function. Weeks disappear into refactors nobody asked for.
The question that actually decides it
Not "which is cheaper", neither is, reliably. The question is: how much of this project is still a question?
- Mostly answered → fixed price is fair, and the certainty is worth the premium.
- Mostly open → fixed price just means you're paying someone to guess, and then arguing about the guess later.
Which is why splitting the engagement usually beats picking one model for all of it.
The split most projects should use
Run discovery as its own small, fixed-price engagement. One or two weeks, a fixed fee, and a written scope as the deliverable.
At the end of it, the project either becomes knowable enough to fix-price properly, or you have a clear-eyed reason to run it as capped T&M with a real backlog. Either way you paid a small, bounded amount to remove the biggest risk in the whole engagement, which is starting a large project on top of an assumption.
Questions worth asking any agency
Whichever model is on the table:
- What specifically happens when scope changes? Get the mechanism, not the reassurance.
- What's the buffer in this number, and what happens to it if the work goes smoothly?
- What am I paying for in week one that I'd still have if we stopped in week two?
- Who owns the code and the accounts if this ends early?
An agency that answers those crisply is telling you something useful about how the whole engagement will run.
Where we land
We quote fixed scope and fixed price after discovery, because by then the guessing is mostly done and the certainty is worth something to both sides. Changes are priced in writing and approved before they're built, which keeps the "that's out of scope" conversation from ever becoming a surprise.
That isn't a moral position. It's just the model that matches how we prefer to be held to account.