What will this project cost?
An estimate without a stated buffer is a promise. This keeps the two apart so you can quote one and plan the other.
Total with contingency
—
Base estimate, before contingency
—
The buffer itself
—
Formula: hours × rate × (1 + contingency ÷ 100)
What the number assumes
A blended rate averages everyone on the project. If seniority varies a lot, calculate the roles separately and add them, because a blend hides which part of the estimate is fragile.
The formula defines the arithmetic. It does not define your business. A figure from this page is a starting point for a decision rather than the decision.
How to estimate what a project will cost
A project estimate is the number of hours you expect it to take, multiplied by the rate those hours cost, plus a contingency you state out loud. The arithmetic is trivial. Everything difficult about estimating is in the first number, and no calculator can help with that. But keeping the buffer separate from the estimate does help, because it stops the two being confused later.
The distinction matters most when things go wrong. If the buffer is a visible line, a scope change is a conversation about the buffer. If it was quietly folded into the hours, the same conversation becomes an argument about whether you estimated badly.
Why estimates are systematically low
People underestimate work they have not started. This is one of the most robustly observed patterns in planning research, and knowing about it does not fix it. Teams that are told about the bias still produce optimistic estimates. It is not a character flaw; it comes from estimating the path you can see rather than the one you will take.
The practical consequence is that a buffer is not padding. It is the correction for a known and predictable bias, and a project quoted without one is quoted below cost more often than not.
The most reliable correction is your own history. If your last five projects came in 30 per cent over, the sixth will too, whatever the estimate says. That ratio is worth more than any general rule of thumb, and it takes one spreadsheet to work out.
Estimating in ranges rather than points
A single number invites a single expectation. Quoting a range, and saying what would put you at each end of it, moves the conversation to the assumptions, which is where it belongs. "160 hours if the brand assets arrive in week one, 210 if they arrive in week three" is a far more useful sentence than "about 180".
Break the estimate down far enough that each piece is something you have done before, and no further. Estimating a fifty-hour task as fifty hours is a guess; estimating it as six things you recognise is arithmetic on top of experience.
What the total does not include
This figure covers labour. It does not cover licences bought for the project, freelancers on a different rate, travel, or the cost of the sales work that won the job. For a studio, business development can be a substantial share of the year, and it has to be recovered from delivered projects, usually inside the rate rather than as a line on the estimate.
It also assumes the hours are actually available. A hundred and sixty hours does not mean four weeks for one person, because nobody delivers forty productive hours a week. That is what the capacity arithmetic is for, and it is a different question from what the project costs.
Questions
How much contingency is reasonable?
It depends on how much of the work is new. Repeat work of a kind you have done before rarely needs more than 10%. Genuinely new work often needs 30% or more, and saying so is more professional than discovering it later.
Should the client see the contingency?
That is a commercial decision, not a mathematical one. Showing it invites a negotiation about the number, and hiding it means overruns look like your mistake.
How large should the contingency be?
It depends on how much of the project you have done before. Work that closely resembles something you have delivered carries perhaps 10 to 15 per cent. Work with an unfamiliar client, an unfamiliar tool or an undefined scope needs 25 to 40 per cent. The honest signal is not the size of the buffer but whether you can say what it is for.
Should I show the contingency to the client?
Show the total, and be able to explain the buffer if you are asked. Hiding it inside an inflated hour count means you cannot defend the estimate line by line, and it means that when the scope changes you have no room left because you have already spent the buffer on nothing.
What is a blended rate and when should I not use one?
A blended rate is a single average rate for a mixed team. It is fine when the mix is roughly what you always use. It breaks down when a project is unusually heavy in one role. A project that is 80 per cent senior design work estimated at the studio's blended rate will lose money, and the loss will look like a delivery problem rather than a pricing one.
Related calculators
Same corner of the arithmetic, different question.