Most briefs are documents. They contain information. They describe what the client wants in terms of what the client can see — the homepage, the navigation, the look and feel. They are written as specifications for a thing that does not exist yet, which means they specify the wrong things.
A good brief is not a document. It is a shared understanding of what success looks like and what the obstacles to it are. The document is evidence that the understanding exists. It is not the understanding itself.
What a brief is not
A list of pages is not a brief. It tells an agency what to build. It does not tell them why. An agency that builds exactly what is listed will produce exactly what was asked for — and frequently something that does not solve the actual problem.
A competitor reference list is not a brief. It tells an agency what you admire. It does not tell them what you need. The agency that builds you the site you admire will give you someone else's solution to someone else's problem.
A visual direction board is not a brief. It tells an agency the aesthetic you are drawn to. Aesthetics are downstream of strategy. A beautiful site that is not structured around the right goals is a beautiful failure.
What the agency actually needs to know
First: who is this for. Not a demographic profile. Not a persona document. A specific description of the person who is going to land on this site with a problem and leave with a solution — or not. What problem do they have. How are they finding the site. What is the one thing they need to understand to take the next step.
Second: what the next step is. Every site has one primary action. Not three. Not a journey. One action that, if more people take it, makes the business better. Everything else on the site exists to support that action or to remove the obstacles to it.
Third: what is currently getting in the way. Why is the existing site not working. What do users do when they arrive. Where do they leave. What questions do they ask that the site does not answer. If there is analytics data, share it. If there are support tickets or sales call recordings, share those. Real user behaviour is worth ten persona documents.
Fourth: what cannot change. Constraints are useful. The logo that cannot be redesigned. The integration that must be preserved. The page that must exist for legal reasons. Knowing the constraints early prevents the agency from designing something that cannot be built.
The conversation the brief should start
A brief should generate questions, not close them. An agency that reads your brief and has nothing to ask has either understood everything immediately — which is unlikely — or has not engaged with it seriously enough to find the gaps.
The useful questions are uncomfortable ones. Why does this page need to exist. What happens if we cut this section. Who is the decision-maker for this and what do they care about. What does a failed project look like from your perspective.
If the agency is not asking these questions, ask them yourself. The answers will improve the brief more than any amount of writing.
What to send on day one
One document. Four sections.
The problem: what is not working and why it matters. The user: who lands on the site, what they need, what they do now. The goal: the one primary action and how you measure it. The constraints: what cannot change and why.
Everything else — the pages, the references, the visual direction — follows from this. If it does not follow from this, it probably should not be in the brief at all.