Most briefs are written once and never opened again. The ones that survive share a trait: they answer the questions engineers ask on day three, not the ones executives ask on day one.
Start with the decision, not the feature
Describe the decision a user is trying to make and what stops them today. Features become easier to cut when everyone can see which decision they serve.
Five questions every brief answers
- Who is this for, specifically?
- What do they do today instead?
- How will we know it worked?
- What is explicitly out of scope?
- What would make us stop?
Keep it to one page
If the brief needs a second page, the scope probably needs a second release. Link out to research and designs rather than pasting them in.
Write down what done looks like
A brief without acceptance criteria invites endless polish. Two or three observable outcomes, agreed before work starts, settle most scope debates in minutes.
Revisit it at every demo
Open the brief at each sprint review and check the work against it. When the brief and the product disagree, update one of them on the spot.