How should features be brainstormed before consulting AI?

How should features be brainstormed before consulting AI?

Brainstorming before consulting AI helps you establish your own understanding of the product. The first list does not need to be complete. Its purpose is to capture observed problems, desired outcomes, and constraints before generated suggestions influence the direction.

Describe users and their current workflow

Identify who experiences the problem and how they handle it today. A wedding organizer may track events in a calendar, tasks in a spreadsheet, and guest responses in message threads.

Walk through the process from beginning to end. Note where information is duplicated, decisions are delayed, or responsibilities become unclear. These observations provide a stronger foundation than copying feature lists from unrelated applications.

Separate what users have actually said from your assumptions.

Convert problems into candidate capabilities

For each difficulty, suggest a capability and explain its intended benefit. Scattered event details might suggest a shared event list. Unclear responsibilities might suggest task assignment and visible status.

Write down dependencies and open questions. Task assignment may require a collaborator model, while event-specific RSVPs require clear invitation rules.

Group related ideas, but avoid choosing every interface detail or technology at this stage. The aim is to explore product behavior.

Ask AI to challenge and extend the list

Give AI the audience, current workflow, candidate features, and constraints. Ask it to identify missing steps, overlapping ideas, edge cases, and assumptions that need user feedback.

Keep a separate list for promising suggestions that fall outside the initial scope. Evaluate each idea by the user outcome it enables, implementation cost, and evidence of need. This preserves human direction while making productive use of AI’s ability to generate alternatives.