How to estimate project hours without lying to yourself

Last updated: July 2026

Ask a freelancer how long a project will take, and they will usually give you an estimate based only on a perfect, trouble-free schedule. The version where the files arrive on time, the feedback is clear, nobody gets sick, and focus is infinite. Psychologists call it the planning fallacy. Clients call it "why is this late".

In short: estimate freelance projects in half-day pieces, add the invisible 15 to 20% of overhead (emails, files, small fixes), and let the buffer cover only true surprises. The free Project Price Calculator turns the hours into a quote.

Estimations cannot be precise, but they can be honest, and being honest is enough. Your buffer handles whatever your estimate cannot predict.

Break it down until there are no hidden surprises

A 40-hour estimate is a guess. Ten 4-hour estimates are a plan.

Decompose the project until no piece is bigger than about half a day. Big blocks hide work inside them. Small blocks force it into the open.

For example, "Design the landing page" hides three review cycles and a mobile pass. In contrast, "Wireframe", "desktop layout", "mobile layout", "revisions round 1", "revisions round 2" hide almost nothing.

Then estimate the pieces, not the project. Most people are bad at estimating long time frames, but surprisingly decent at predicting a single afternoon's work.

Count the work that never makes the task list

Every project carries hours that appear on no brief. The kickoff call, writing up feedback, chasing missing assets, exporting and delivering files, the countless short emails. This is real work done for this client on this project, and it vanishes from every self-made estimate I've ever seen, including mine.

Add them as their own pieces. If that feels like inflating the estimate, notice that the hours will happen either way. The only question is whether they get priced.

Estimate vs buffer: two different jobs

A common mistake is using the buffer to cover up a weak estimate. The two are meant to protect you against different things.

The estimate covers the work you can already see: the individual tasks, plus hidden overhead like project management and communication. The buffer covers what nobody can see yet: the surprise stakeholder, a third round of revisions, or a file that arrives broken. If you find yourself thinking "the buffer will absorb the emails," remember that the time spent on emails belongs in your main estimate.

Conversely, padding every single task and keeping a large buffer will make your prices too high, causing you to lose jobs you could have won.

An estimate is a promise you make to your own calendar.

Type the honest total

Add up your tasks, type them into the tool, and add a 15% to 25% buffer. The final price you see is the safe quote, the one that already covers the cost of late files and delayed feedback.

The shortcut that beats every method

Your past projects. Actual hours, if you tracked them; honest recollection, if you didn't. A single note after each delivery ("estimated 30, took 41, the extra was revisions") builds a private reference base that beats any formula within five projects. The planning fallacy survives on forgetting. Write things down and fix it.