How long a job actually takes: estimating properly

The three costs nobody prices, the difference between working time and lead time, and the three-month habit that fixes the problem for good.

Published on September 14, 2026Written by Kalepio7 minute read

You said three days. It took seven. And you were not slow: you worked normally, on a normal job, and the gap is still more than double.

This is not an optimism problem. It is that you priced the work, and the job is not only the work.

The three costs nobody prices

Getting back into it

The biggest and the most invisible.

Every restart after an interruption costs an hour to an hour and a half: reopening the files, finding where you were, rereading the brief, getting the project back into your head. The change itself takes three minutes; the getting-back-in is constant and does not depend on the size of what you came to do.

A job spread over three weeks has eight or ten of them. That is where most of the gap comes from, and it is the same arithmetic as in how many rounds of feedback to plan for: fragmented work costs double the same work done in one block.

The job's own admin

The brief to write, the quote, the emails, the chasing, the invoice, the delivery, filing the files.

Count fifteen to twenty-five per cent on a small job. And more importantly: that cost barely shrinks with the size of the project. A three-hundred-pound job carries roughly the same admin as a three-thousand-pound one.

That is the complete explanation of something many people live without naming: small jobs are not profitable at the same hourly rate. It is not an impression, it is a fixed cost divided by a smaller fee.

The client's own time

Not your time, which is why it gets forgotten. But it is calendar time, and a promise is made in calendar time.

Three days of work spread across a job where the client takes four days to answer each question comes to three weeks. Your estimate of the work was right; your lead time was wrong; and the lead time is what the client remembers.

The distinction that fixes half the problem

Working time and lead time are two different numbers, and you have to state both.

"Three days" is a trap you set for yourself, because the client hears "Thursday". What to write is the double sentence:

Three days of work, delivered within two weeks of receiving your copy.

Three elements, and the third is the one people leave out: the lead time starts from an event, not from signature. A lead time running from the signed quote is a lead time the client can consume themselves without realising.

How to estimate: decompose, never guess

One rule, and it is enough: never estimate a job, estimate its pieces.

Break it down until you have blocks of half a day at most. And here is the real value of the exercise, which is not the accuracy of the sum: anything you cannot break into half-days is a piece you do not yet understand. That is precisely the one that will blow up.

When you hit such a block you have three honest options and one bad one. The honest ones: ask the client a question, take it out of scope, or price it separately. The bad one is putting a comfortable number on it and hoping.

The three-month habit

Here is the most useful thing in this article, and almost nobody does it because it feels like an admission.

For three months, on every job, write down two numbers: what you estimated, and what it actually took. Two columns, nothing more.

After three months you will have your ratio, and two things will surprise you. The first is that it is stable: it does not vary much from job to job, because what causes the gap is structural rather than accidental. The second is that it sits between 1.4 and 2 for most people.

Then you multiply. That is all. You do not need to estimate better, you need to know your own gap, which is far easier and can be measured instead of improved.

Never pad in silence

The temptation, when a piece is vague, is to quietly inflate the whole thing.

That has two effects and both are bad. You become expensive on the parts you know inside out, which is exactly where a client compares. And you still have not dealt with the uncertainty: it is hidden, so it is neither discussed nor billed, and if it materialises you are the one who pays.

Say it instead. "The number of pages will be fixed after the scoping session, each additional page is at this rate" is a sentence that protects everybody, and a client understands it immediately. On what belongs in the document: what a freelance quote needs.

When you are already late

One rule: say so the moment you know, not at the deadline.

You know on Tuesday that Thursday will not hold. A message on Tuesday is information; the same message on Thursday evening is bad news plus a disappointment. It is the same mechanism as in what your client cannot see: what damages trust is almost never the delay, it is having been left with no news until the deadline.

And give a new date, a specific one. A delay with no new date is a delay with no end.

Frequently asked questions

Why does a job always take longer than planned?

Because people price the work rather than the job. Three costs escape the estimate: getting back into it after each interruption, an hour to an hour and a half each time, the job's own admin at fifteen to twenty-five per cent, and waiting for the client's answers.

How do you estimate a freelance project?

By breaking it into blocks of half a day at most. The value is not the accuracy of the sum: anything you cannot break down is a piece you do not yet understand, and that is the one that will blow up.

What is the difference between working time and lead time?

Working time is what it takes you, lead time is what the client waits. Three days of work on a job where they take four days to answer each question comes to three weeks. State both, and start the lead time from a specific event.

Should you pad your estimates to be safe?

Not in silence. You become expensive on the parts you have mastered, which is where a client compares, and the uncertainty stays hidden so it is never billed. Say it: "the number of pages will be fixed after scoping, each extra page is at this rate".

Why are small jobs less profitable?

Because a job's admin barely shrinks with its size. A three-hundred-pound job carries roughly the same brief, the same emails and the same invoice as a three-thousand-pound one.

Kalepio asks these questions for you

You send one link, your client answers once, and you start with everything you need.

Read next