How to write an MVP brief before you build
A minimum viable product is easier to scope when you describe a complete task instead of a collection of screens. A dashboard, login page, and notification system might all be necessary, but none explains why someone would use the product.
Start with a brief that makes the important decisions visible. It should tell a builder what a user needs to accomplish, what must work on day one, and what evidence would justify another iteration. A short, specific brief is more useful than a long feature list with no priorities.
Choose one user and one outcome
Name the person using the first version and the moment they need it. ‘Students’ is a broad audience. ‘A student recording a shared purchase immediately after paying for it’ describes a situation you can design and test.
Then write the outcome without mentioning an interface: ‘The group can see who owes what.’ This makes it easier to compare different solutions. A simple shared form might test the idea before you need a sophisticated account system.
Map a complete task, including the awkward parts
Write the steps from the first input to the useful result. For a shared-expense product, that could mean creating a group, adding a purchase, choosing participants, and seeing updated balances. Treat this as a proposed workflow, not a description of any particular product's full feature set.
Include what happens when a person makes a mistake. Can they correct an amount? What if two people record the same purchase? These questions often expose more important requirements than another dashboard widget.
- Entry: what brings the user here, and what information do they already have?
- Action: what is the smallest sequence that produces the result?
- Recovery: how can they correct, cancel, or retry the task?
- Completion: what visible evidence tells them it worked?
Separate requirements from later possibilities
Divide the work into three groups: necessary for the core task, necessary for safe operation, and possible later improvements. Authentication, permissions, data recovery, and clear failure messages can belong in the second group even when they are not the feature that attracts users.
State the constraints that change the design: mobile use, unreliable connectivity, an existing database, language requirements, or a fixed launch window. If an external service is essential, identify who owns the account and what the product should do if that service is unavailable.
Keep a visible list of deferred ideas. This prevents every discussion from reopening the same scope decisions, while preserving useful possibilities for a later release.
Define evidence before you define success
Choose a measure tied to the user outcome. For an expense tool, ‘a group can record and settle a week's shared purchases without a separate spreadsheet’ is more informative than counting account registrations alone.
Pair a measurable check with a conversation. Observe where a small group gets stuck, what they still do elsewhere, and whether they come back without prompting. Record the baseline and the observation window; otherwise a later percentage has little context.
A brief you can use
Use these prompts as a starting document. Answer in plain language, attach examples of the current workflow, and mark assumptions that still need testing. The goal is shared understanding before implementation decisions become expensive to change.
- User and situation: who needs this, and when?
- Current approach: how is the task handled today, and where does it fail?
- Core outcome: what must a person be able to finish?
- First workflow: what are the steps, inputs, and visible result?
- Operational essentials: permissions, corrections, failure handling, and data ownership.
- Constraints and exclusions: what shapes the first release, and what is deferred?
- Validation: what will you observe, over what period, before deciding what to build next?
See the products behind these notes in our work, or start a conversation about a project of your own.