How to Write a Design Brief That Gets You What You Wanted
Describe the problem, not the solution.
Published
Topic
Design
Almost every project that goes wrong went wrong at the brief. Not at the design, not at the revisions, not at the personalities. At the twenty minutes at the start where nobody wrote down what the actual problem was.
I have been on the receiving end of several hundred briefs. Here is what separates the ones that produce good work from the ones that produce four rounds of revisions and a tired client.
Why briefs fail
There is one failure mode and it accounts for most of them.
The brief describes the solution instead of the problem.
"Make it modern and minimal. Use blue. Something like Stripe but warmer. Sans-serif. Maybe an abstract mark."
Every sentence there is an instruction about output. None of them tells me anything about the business, and all of them are guesses at a solution made by someone who has explicitly hired a specialist because they are not the specialist.
The tragedy is that these briefs feel helpful to write. They feel specific and decisive. What they actually do is constrain the designer to your first instinct, which is almost never the best available answer, and then leave everyone confused about why the result feels flat.
The other reliable failure is the opposite extreme. "Surprise me, you're the creative." That is not freedom, it is an absence of information, and it guarantees a first round that misses because there was nothing to aim at.
What a designer actually needs to know
In rough order of usefulness.
What the business does, in one plain sentence. No mission statement, no adjectives. What you sell and to whom.
Who buys it. Specifically. Not "everyone" and not "25 to 45 year old urban professionals." Describe an actual person who bought something last month.
Who you are compared against. Name five competitors and send links. This single item does more work than any other line in a brief, because it tells me what your customer is seeing right before they see you, and distinctiveness is measured against that set rather than in the abstract.
What has changed. Why now? A launch, a raise, a shift in what you sell, a new market. The trigger explains the urgency and usually reveals the real problem. If nothing has changed, that is worth knowing too, and it is worth reading When Does a Business Actually Need a Rebrand before spending anything.
What success looks like. How will you know this worked? "It looks better" is not measurable and cannot be argued about productively. "A founder can send this deck without apologizing for it" can.
The constraints. Budget, deadline, where this has to work, languages, accessibility requirements, existing assets that cannot change, technical limits. Constraints are not the boring part of a brief. They are the part that makes the answer findable.
Who decides. One name. Not a committee. If five people have veto power, say so upfront so the process can account for it, but understand that it will cost you distinctiveness for the reasons I set out elsewhere.
What you have already tried, and why it did not work. Enormously useful and almost never included.
References, and how to use them properly
Sending references is good. Sending references without explanation is nearly useless.
"I like this" tells me nothing, because I cannot tell which part you like. The layout? The color? The photography? The tone? The type? You might be responding to the copywriting and not know it.
The fix is one sentence per reference:
Bad: a link to a website.
Good: "I like how restrained this is. There is almost nothing on the page and it still feels expensive."
Also good: "I do not like this, and I am including it because it is what everyone in our category looks like."
Negative references are worth as much as positive ones and hardly anyone includes them. If there is a specific thing you never want to see, say it now rather than in round three.
What to leave out
Solution language. Colors, typefaces, mark styles, layout instructions. If you have a genuine constraint, an existing brand color you must keep, say it as a constraint. Otherwise, hold it.
Adjective piles. "Modern, clean, professional, innovative, trustworthy, bold but approachable." Every brief contains this list, it is the same list every time, and it eliminates nothing. If your brief's adjectives would also be true of your competitors, they are not doing any work.
Internal politics dressed as requirements. If a board member insists on green, say that plainly. It is a real constraint and it can be designed around. Disguising it as a strategic preference wastes everyone's time.
Everything. A forty page brief is not thorough, it is unfiltered. Two pages that answer the questions above beat forty that bury them.
A brief structure you can copy
Roughly two pages. That is enough.
The business. One paragraph. What you do, who for.
The trigger. Why this project, why now.
The problem. What is failing today, described as a problem rather than as a missing deliverable. One paragraph.
The audience. Who needs to react to this, and what you want them to think.
The competitive set. Five links, one line each on how you differ.
What success looks like. Two or three specific outcomes.
Scope. What is being made. Be explicit about what is not.
Constraints. Budget, deadline, technical, legal, existing assets.
References. Five to eight, each with one sentence on what you are responding to. Include at least one negative.
Decision maker and process. Who signs off, how many rounds, how feedback will be delivered.
If you cannot fill in point 3 without describing a deliverable, stop. That is the sign that the problem has not been diagnosed yet, and it is the single most valuable thing you can discover before spending money.
The brief is only half of it
A good brief gets round one close. Feedback gets you the rest of the way, and it fails in exactly the same way briefs do.
Describe the problem, not the fix. "Make the logo bigger" is a solution. What you usually mean is "it does not feel like the most important thing on the page," and there are six ways to solve that, of which making the logo bigger is often the worst.
Consolidate. One voice, one document, one round. Six people sending contradictory notes separately produces work that satisfies nobody, because the designer is forced to average opinions instead of resolving them.
Separate taste from function. "I do not like green" and "our largest customer's brand is green" are different notes with different weights. Say which one you are giving.
Answer the question that was asked. If you were sent three directions to choose between, choose one. Combining elements from all three is how you get a design with no argument holding it together.
Be timely. Feedback that arrives a week late does not just delay the project by a week. It costs the momentum and the context, and the work restarts colder than it stopped.
The short version
Tell your designer what is broken, who it is broken for, and what you are up against. Tell them what you cannot change and what happens if this works.
Then let them solve it. That is the thing you are paying for, and a brief full of your own solutions is an expensive way to hire someone and then do their job badly.