Proposals

How to scope a project from a client call

Four pages of notes into a scope somebody could hold you to. The step everyone skips is the translation — turning what they said into things that can be counted and finished.

N

North

2026-06-08 · 4 min read

How to scope a project from a client call

You have four pages of notes and a client expecting a document. The distance between those two is where most of the dread in freelancing lives, and it is not a writing problem. It is a translation problem, and it goes wrong in a specific place.

Here is the direct answer to the question. Take each thing the client wants, restate it as an object with a quantity and a finish line, list what you had to assume to do that, and put both in the proposal. Everything below is how.

Translate ambitions into countable things

Clients speak in outcomes. "We need the brand to feel more grown up." "The site should actually sell." Those are correct statements of what they want and completely useless as scope, because neither can be finished.

So convert. Not "refresh the brand" but: a revised wordmark, two lockups, a colour and type system, and a six-page usage sheet. Not "make the site sell" but: five page templates, a rewritten homepage, and a checkout flow rebuilt to three screens.

The test to apply to each line is whether the two of you could look at it in eight weeks and agree it was done without needing to discuss it. If a discussion would be required, the line is not finished being written.

This is also the step where you find out what you do not know. You cannot say "five page templates" until you know there are five pages, which means the vague version was hiding a question you now have to ask. Good. Ask it before you quote rather than after.

Write the assumptions down as assumptions

This is the part almost nobody does and it is the highest-value paragraph on this page.

Every estimate rests on things you decided were probably true. The copy will be supplied. There is one round of stakeholder review, not three. The photography exists. The developer will be available in April. You made all of those calls silently while adding up the days.

Put them in the document as a numbered list under a heading that says Assumptions. Not buried, not softened. When one turns out to be wrong — and one always does — the conversation is about a fact that changed rather than about whose memory is better. That is a different conversation and it is one you can win politely.

There is an illustration of the alternative in a Hacker News thread that anyone who has been through it will recognise. A freelancer's client decided, after the work was delivered, that each task should have taken about two hours. One of them had taken seven. The client refused to pay the difference. Note what had happened: the client had been carrying a private estimate the whole time, never said it, and only revealed it at the invoice. The thread's verdict was blunt, and correct, that the client was trying to stiff him and it would only get worse. But the structural lesson sits earlier than that. There were two estimates on that project and only one of them was ever written down.

Keep their words, use your nouns

Two things that look contradictory and are not.

The client's language belongs in the framing. If they said "rebrand," the opening paragraph says rebrand, because a document that renames their problem reads as though it was written for somebody else.

Your language belongs in the deliverables. This is where precision beats rapport. "Brand refresh" is not a deliverable in any vocabulary, and softening a specific noun to match their casual phrasing is how you end up owing three logo directions you never agreed to.

Then the boundaries, and the price

Two more lists and you are done. What is explicitly not included, named specifically rather than generically, which is the whole apparatus for keeping scope from drifting. And the phases, so the money has somewhere natural to attach.

Then price the outcome rather than the hours — the method for that is a longer subject, and the short version is that your scope list tells you what the work costs you while only the client can tell you what it is worth to them. You need both numbers and they come from different places.

North does this translation from your notes: deliverables with quantities, an exclusions list, phases, and the priced options, as a draft you correct rather than a page you start.

The thing worth remembering, if you remember one thing: the argument you are going to have in week six is already latent in the sentence you are writing now. Write the sentence so it cannot be read two ways, and there is no argument to have.

Share this post

Frequently asked questions

How do I scope a project from a client call?
Translate every ambition into a countable deliverable, then write down what you assumed to get there. The assumptions are the part people leave out, and they are what turns a disagreement in week six into a change order rather than an argument, because a written assumption that turns out to be wrong is visibly a new fact rather than a difference of memory.
What should I write down during a discovery call?
Their exact phrasing, the numbers they say out loud, and every noun that sounds like a deliverable, including the ones dropped in passing. Record the numbers even when they seem irrelevant — how many products, how many locations, how many people need training — because those are what convert a vague ambition into a quantity you can price.
How detailed should a scope be before I send a proposal?
Detailed enough that both of you would agree, without discussion, whether each item is finished. That is a higher bar than it sounds: 'website' fails it, 'five page templates built in Webflow on your hosting' passes it. Anything you cannot state that clearly is not scope yet, it is an intention.
What if the client changes their mind after the scope is agreed?
They will, and a well-written scope makes that a priced conversation rather than a fight. What makes it work is the exclusions list and a named number of revision rounds; what breaks it is agreeing verbally to something small in week two and never writing it down, because the second small thing then has no line to be measured against.

From the blog

Plain-spoken guidance on the business side of your craft.

View all posts