How might we questions framed from user research findings
Turns research findings into well-framed problem statements and "how might we" questions at the right scope, neither so broad they mean nothing nor so narrow they already contain the solution, then picks the best three to take into ideation. Use it after interviews, usability tests or support analysis, before a brainstorm or design sprint.
The output this prompt produced, unedited.
The prompt
Fill in these first
Product and user group: [e.g. a learning app, parents of students in classes 6 to 8]
Research done: [e.g. 12 parent interviews and 300 app reviews]
Key findings, each with the evidence behind it: [paste, e.g. "parents cannot tell if practice time was real study, 9 of 12 mentioned it"]
Goal or metric the team is working on: [e.g. weekly active parents]
Constraints: [e.g. no new hardware, launch within a quarter]
Do this
Rewrite each finding as an insight: what users do, why, and the tension behind it.
Write a problem statement for each: "[User] needs a way to [need] because [insight]."
Write three how might we questions per problem:
Broad: opens many directions
Balanced: focused but leaves the solution open
Narrow: close to a solution, used only as a check
Remove any question that names a specific feature or solution.
Output format
| Finding | Evidence | Insight | Problem statement |
| Problem | Broad | Balanced | Narrow |
The three balanced questions to take into ideation, each with why it matters for the goal, and two example ideas only to show the range.
Guardrails
Every insight must point back to evidence in the inputs; no invented user quotes.
Avoid "how might we make users" phrasing that frames people as the problem.
how might we questionsproblem framingdesign thinkingproblem statement from researchideation promptsux research synthesis
How to run this
Copy the prompt and paste it into Gemini, then set the aspect ratio to 3:4. No reference photo is needed for this one.
The prompt is written in Markdown with each part under its own heading, which is what the copy action hands over. Change one section at a time, keeping the rest intact, so you can tell which change did the work.