Fixed price or hourly is usually argued as a preference. It is not. It is a decision about who carries the risk of what nobody knows yet, and the honest answer changes from project to project.
The question under the question
On a fixed price, the agency carries the risk. If the work takes longer than estimated, the agency pays for the difference in margin. On time and materials, the client carries it: they pay for the hours the work actually takes, and they accept that the final figure is not known on day one.
Neither is generous or greedy. What matters is whether the risk sits with the party that can actually control it.
When a fixed price is the right call
- You have done this shape of work before, and you can look up what the last three similar projects really took rather than what you remember.
- The scope can be written down as deliverables, rounds of revision and an explicit list of what is out of scope.
- There is a change process both sides accepted before the work started, so the first new request does not become an argument.
- The client needs a number to approve a budget. Sometimes the fixed price is not about risk at all; it is about how their finance department works.
When hourly is the honest answer
- Discovery. If the first phase exists to find out what the project is, no one can price the whole of it yet.
- A moving target. A client in the middle of a rebrand, a funding round or a reorganisation will change direction, and a fixed price only hides that until the invoice.
- Ongoing work. Support, maintenance and retainers are priced by capacity, not by a deliverable.
- A first project with a new client. You do not yet know how many rounds of feedback they need. That is not a flaw, it is missing information.
The version most agencies end up with
A fixed price for a well-defined phase, then time and materials for what follows. Discovery hourly, build fixed. Or a fixed monthly capacity with a written rule for what happens when a month goes over. The point is that the pricing model matches how much is actually known at that moment, rather than being a company-wide rule applied to every project.
Pricing a fixed bid without guessing
- Estimate in hours by role, not in one number. Design, build and project management have different cost rates.
- Check the estimate against the past. What did the last similar project actually record? If you cannot answer that, the estimate is a wish, and the case for time and materials just got stronger.
- Add contingency openly. A percentage for the unknown, decided before the client sees the number, not discovered afterwards.
- Price against cost, then against value. Cost tells you the floor; value tells you whether the ceiling is higher.
- Write down what is out of scope. The out list protects the margin more than the in list does.
What the numbers do
Take a website quoted at 8,000. The estimate is 120 hours at a 35 cost rate, so 4,200 of effort, plus 1,000 of direct expenses. That leaves 2,800 before company overhead, a 35% margin.
Now the same project takes 160 hours because the client needed two more rounds: 5,600 of effort plus the same 1,000 leaves 1,400, or 17.5%. The price did not change. The risk did what risk does.
On time and materials the same 40 extra hours would have been invoiced, and the conversation would have happened while the work was in front of both parties instead of at the end.
The change process matters more than the model
Whichever model you choose, extra work needs a route: a request with one owner, a decision that says included, extra or declined, and a record kept with the project. Without it a fixed price quietly loses margin and an hourly project quietly loses the client’s trust. There is more on that in scope creep and billing.
Where PLYNTUM fits
requests and approvals gives every change an owner and a review status. time tracking records the hours against the project, which is what makes the next estimate better than the last one. finance and costs puts effort, expenses and revenue on one record, so you can see which model actually worked on which kind of project.