Loading
August 5, 2026

How to Brief a Designer So You Actually Get What You Wanted

Most design revisions are caused by the brief, not the designer. What to include, what to leave out, and how to give feedback that produces a better result.

When a design project goes through seven rounds of revision, the cause is almost never the designer. It is the brief, and specifically the parts of it that were in someone head rather than on paper.

A good brief takes twenty minutes to write and saves weeks. Here is what belongs in one.

What the design has to achieve

Start with the job, not the artefact. “A poster” is not a brief. “A poster that gets people walking past our shop to know we now do home delivery” is.

Every design decision downstream depends on this. Whether the delivery message or the brand name is largest, whether the phone number needs to be readable from across the road, whether it should look premium or approachable. Without the objective, the designer is guessing at all of it.

Who is looking at it, and where

Audience and context change everything. A hoarding read at forty kilometres an hour and an Instagram post examined at arm length are entirely different problems, even with the same message.

Specify: who this is for, where they will see it, how long they will look, and what they should do next. A designer who knows the piece will be seen for two seconds will make radically different choices about how much text to include.

The one thing it must communicate

If everything is important, nothing is emphasised. Force yourself to name the single most important message, then rank the rest.

This is the most useful sentence in any brief and the one clients most resist writing, because it means accepting that some information will be small. But a design where five things compete for attention communicates none of them.

Exact copy, checked

Supply final text, spelled correctly, with names and numbers verified. Not approximate text to be finalised later.

Copy changes after layout mean the layout has to change too, and text length affects design in ways that are hard to see until you try. “We will confirm the phone number later” is how a hoarding goes up with the wrong number on it.

Practical constraints

Everything that limits the solution should be stated upfront rather than discovered at revision three.

  • Exact dimensions, or where the publication size will come from
  • Print or screen, and if print, which process
  • Mandatory elements: logos, legal text, certifications, partner marks
  • Brand colours and fonts if they exist
  • Anything explicitly forbidden
  • The actual deadline, and what happens after it

References, with reasons

Showing examples helps enormously, but only with an explanation. “I like this” tells a designer nothing. “I like this because the offer is instantly readable and it does not feel cheap” tells them what to reproduce.

Include negative references too. Knowing what you dislike and why narrows the space faster than positive references alone.

One caution: references should guide direction rather than be copied. Asking for a version of a competitor design produces something that makes you forgettable in the one place where being distinctive matters.

Who approves

Name one decision-maker.

Design by committee reliably produces the least interesting option, because it is the only one nobody objects to. Collect input from whoever needs to be consulted, then have one person consolidate it into a single set of instructions.

The alternative is contradictory feedback from four people, which the designer cannot satisfy and will not try to.

Giving feedback that helps

The most common feedback failure is prescribing a solution instead of describing a problem.

“Make the logo bigger” is a solution. The underlying problem might be that the brand is not registering, and the actual fix might be more white space rather than a larger logo. Describe what is wrong and let the designer solve it; that is what they are for.

Useful feedback sounds like:

  • “The offer is not the first thing I notice, and it should be”
  • “This feels formal, and our customers are families with young children”
  • “I cannot read the phone number on my phone screen”
  • “The green does not match the green in our other material”

Unhelpful feedback sounds like:

  • “Make it pop”
  • “I do not like it, try something else”
  • “Can we see a few more options”
  • “My wife thinks it should be blue”

That last one is not a joke. Outside opinion collected casually is the single most common cause of a project losing direction, because the person offering it has none of the context and no responsibility for the outcome.

Consolidate feedback into one round

Sending three separate messages over two days, each with new thoughts, produces three partial revisions and confusion about which instruction is current.

Sit with the design, collect everything, and send it once. Number the points. It is faster for everyone and the result is better.

Agree revision rounds upfront

Most design quotes include a defined number of revision rounds, and this protects both sides. It gives you clarity on what is included and gives the designer a reason to get it right early rather than iterating indefinitely.

If a project genuinely needs more rounds because the requirement changed, that is a scope conversation rather than an argument.

Ask for source files

At the end, get the editable source files along with the exports. It costs nothing to agree at the start and means you are never dependent on one supplier to make a small change two years later.

An agency reluctant to hand over source files is protecting a dependency rather than a standard.

A brief template you can copy

  • What this is: the piece and where it will appear
  • Objective: what should happen because of it
  • Audience: who, and in what context
  • Primary message: the one thing, then everything else ranked
  • Copy: final, checked text
  • Must include: logos, legal, mandatory elements
  • Constraints: size, format, colours, anything forbidden
  • References: two or three, with reasons
  • Approver: one name
  • Deadline: the real one

Twenty minutes on that saves a fortnight of revisions, and produces work you actually wanted rather than work you eventually accepted.

See how we run branding and design projects, or tell us what you need designed.