Every agency proposal is really a decision about who carries the risk. Hourly billing puts it on the client: if the work takes longer, you pay more. Fixed price puts it on the agency: if the work takes longer, they absorb it. Both models are legitimate. Both are routinely used to hide a bad deal. After six years of pricing projects — and inheriting a good number that went wrong elsewhere — here is how I would choose if I were on the buying side.
What hourly actually does
Hourly feels transparent. You see the timesheet; you pay for what was done. Three things undermine that in practice.
- The estimate is not a price. "Around 120 hours" is a guess made before anyone has opened your CMS. Web projects have a well-documented tendency to overrun, and on hourly terms every overrun is invoiced. Finishing 30 to 60 percent over the estimate is normal, not exceptional.
- The incentive points the wrong way. Efficiency costs the agency money. Nobody is dishonest about it; it is simply that no one is rewarded for finding the two-hour solution to a ten-hour problem.
- Decisions become billable. The clarification call, the extra revision, the "quick look" at an unrelated bug — each is reasonable, each is an hour, and they add up to the overrun.
Where hourly is right: work that genuinely cannot be scoped. A rescue project on someone else's codebase. Ongoing maintenance. An experiment where the point is to learn. If the end state is unknown, paying for time is honest.
What fixed price actually does
Fixed price feels safe: one number, agreed up front. Its failure modes are quieter.
- The scope is vague, so the price is meaningless. "Website, up to 10 pages" says nothing about integrations, content, revisions or what happens when page eleven is needed. A single number with no scope attached is not a fixed price; it is the opening bid in a negotiation you will have later.
- Everything outside the scope becomes a change order — sometimes reasonably, sometimes as the real revenue model. Watch for a low headline price and an hourly rate for "extras" that turns out to cover most of what you wanted.
- Quality is the release valve. When a fixed-price project overruns, the agency loses money on every additional hour. The temptation is to finish faster: skip the device testing, skip the performance work, ship the template as-is.
Where fixed price is right: anything with a definable end state and a client willing to define it. Which is most websites, most stores, most brand projects and most CRM implementations.
The real comparison
| Hourly | Fixed price (done well) | |
|---|---|---|
| Who carries overrun risk | Client | Agency |
| Incentive for efficiency | Weak | Strong |
| Incentive to cut corners under pressure | Weak | Present — mitigated by acceptance criteria |
| Budget certainty | Low | High, if scope is specific |
| Flexibility mid-project | High | Via change orders |
| Best for | Open-ended, unscoped work | Defined deliverables |
Five clauses that make a fixed price worth signing
The protection is not in the model. It is in what the document says. A fixed-price quote is only safe with these in it.
1. A line-item scope
Pages by name. Features by name. Integrations by name. Who writes the content. Which browsers and devices are tested. If a thing you care about is not on the list, it is not in the price — so get it on the list.
2. Stated exclusions
The honest agency tells you what is not included: third-party licence costs, stock photography, migrating 400 blog posts, hosting. An exclusions section is a sign the scope was thought about; its absence is a sign the change orders are coming.
3. Revision rounds and a change-order process
How many rounds at each stage. What happens on round four. How a scope change is priced and approved — in writing, before work starts, with the timeline impact. This clause protects both sides from the resentment that "unlimited revisions" reliably produces.
4. Milestone payments tied to deliverables
A deposit, then payments on approved design, on staging sign-off, on launch — tied to things you can see, not to calendar dates. You should never be more than one milestone ahead of the work.
5. Measurable acceptance criteria
This is the one almost nobody includes and the one that neutralises the quality risk. "Done" should be a number: Largest Contentful Paint under 2.5 seconds on a mid-range phone, Interaction to Next Paint under 200 ms, zero accessibility errors at WCAG AA, every old URL redirected. If a fixed-price agency will not commit to measurable targets, ask why. If they will, the incentive to cut corners disappears — cutting the corner means not getting paid.
Questions to ask either way
- What is not included in this number?
- What were your last three projects' final costs against their original quotes?
- What do I own at the end — code, files, accounts, domain?
- What happens if the launched site fails the performance targets?
- Who, by name, is doing the work — and will I talk to them directly?
A good agency on either model answers all five without flinching.
What we do, and why
We price projects fixed, with a written scope, stated exclusions, staged payments and a performance budget in the acceptance criteria — and we price ongoing work (SEO, care plans, GoHighLevel retainers) monthly with no minimum term, because that work is open-ended by nature. The fixed number is set after a free audit, not before it, because a price given before anyone has looked at your site is a guess in either model. You can see the ranges on our pricing page; the number in your proposal will be one of them, not a range.
Hourly puts the risk on you and calls it transparency. A vague fixed price puts the risk on you and calls it certainty. A specific fixed price with measurable acceptance criteria is the only version where the agency is paid for the outcome you wanted — which is the only thing you were trying to buy.
Frequently asked questions
Is fixed-price web design more expensive than hourly?
Usually a little, on paper, because the agency prices in the risk of scope surprises. In practice hourly projects routinely finish 30 to 60 percent over the original estimate, so the fixed number is often the cheaper one by the end. The comparison that matters is final cost, not the first quote.
What should a fixed-price web design contract include?
A line-item scope with page counts and features, the exclusions stated explicitly, the revision rounds included, a change-order process with pricing, milestone payments tied to deliverables, the acceptance criteria — ideally measurable ones like Core Web Vitals targets — and a statement that you own the code and accounts on final payment.
When is hourly billing the right choice for a website?
For genuinely open-ended work: ongoing maintenance, experiments, research, or a rescue project where nobody can scope the problem until they are inside the code. For anything with a definable end state — a site, a store, a brand — fixed price protects the client better.
