How Might We 💬 Prompts

Purpose: To reframe user problems or insights into open-ended design opportunities that encourage creative exploration. 

Design Thinking Phase: Define 

Time: 45–60 min session + 1–2 hours analysis 

Difficulty: ⭐⭐ 

When to use:    After synthesising research into clear findings   To align cross-functional teams around actionable problem statements   Before ideation or sprint planning

What it is

“How Might We” (HMW) is a problem framing technique used to translate user insights into structured prompts for ideation. It reframes challenges as opportunities by asking, “How might we…” followed by a targeted problem or potential.

📺 Video by DAN Innovation Council and BOI. Embedded for educational reference.

Why it matters

HMW questions enable product teams to shift from problem-focused mindsets to solution-focused thinking. By framing insights in an open but bounded way, teams can spark collaborative ideation without jumping to solutions too early. It aligns everyone—from stakeholders to engineers—around purposeful design problems.

When to use

•  Right after affinity mapping or thematic research synthesis
•  When reframing vague stakeholder problems into user-centred opportunity areas
•  Before ideation workshops, sprints, or co-design sessions

Benefits

•  Rich Insights: Helps uncover user needs that aren’t visible in metrics.
•  Flexibility: Works across various project types and timelines.
•  User Empathy: Deepens understanding of behaviours and motivations.

How to use it

1.  Gather key user insights or pain points from research (e.g., interview quotes, usability data).
2.  Identify the core challenge behind each insight — what problem is it pointing to?
3.  Reframe that problem into a question using “How might we…” language. Keep it optimistic and open-ended, but specific enough to tackle in ideation.
4.  Write multiple variations if needed — different angles can lead to different solution spaces.
5.  Cluster or prioritise HMW questions based on feasibility, impact, or strategic fit before moving into ideation.

Example Output

From the insight “Users abandon the checkout when shipping options are unclear,” we might generate:

•  How might we help users feel more confident about shipping timelines?
•  How might we make delivery expectations clearer during checkout?
•  How might we reduce uncertainty around shipping costs early in the journey?

Common Pitfalls

•  Too vague: “How might we improve the experience?” lacks direction and won’t guide ideation effectively.
•  Too narrow: “How might we add a chatbot?” forces a specific solution, limiting creative options.
•  Not rooted in research: Skipping the insight phase can result in poorly framed or irrelevant questions.

10 Design-Ready AI Prompts for How Might We – UX/UI Edition

How These Prompts Work (C.S.I.R. Framework)

Each of the templates below follows the C.S.I.R. method — a proven structure for writing clear, effective prompts that get better results from ChatGPT, Claude, Copilot, or any other LLM.

C.S.I.R. stands for:

•  Context: Who you are and the UX situation you're working in
•  Specific Info: Key design inputs, tasks, or constraints the AI should consider
•  Intent: What you want the AI to help you achieve
•  Response Format: The structure or format you want the AI to return (e.g. checklist, table, journey map)

 Level up your career with smarter AI prompts.    Get templates used by UX leaders — no guesswork, just results.   Design faster, research smarter, and ship with confidence.   First one’s free. Unlock all 10 by becoming a member.

Prompt Template 1: “Generate How Might We questions from a research insight”

Generate How Might We questions from a research insight

Context: You are a Product Designer synthesising findings from a usability study on [checkout experience].  
Specific Info: One key insight shows that users [abandon the process] when they encounter [surprise costs at the final step].  
Intent: Transform this insight into a variety of HMW prompts suitable for a design sprint.  
Response Format: Return a list of 5–7 “How might we…” questions with a note on what aspect each one frames (e.g., clarity, trust, timing).

Ask for clarification if the insight lacks user behaviour detail.
Offer one idea for reframing the insight if better HMW questions could emerge.

Prompt Template 2: “Synthesise HMW questions from an affinity map”

Synthesise HMW questions from an affinity map

Context: You are a UX Lead preparing for a team ideation session.
Specific Info: You’ve identified 3 themes from qualitative research: [confusion around sign-up options], [lack of feedback during form submission], and [fear of sharing personal data].
Intent: Generate 2–3 HMW questions per theme to guide the brainstorming workshop.
Response Format: Output a table with themes in column 1 and 2–3 How Might We prompts per theme in column 2.

If any theme feels too broad, suggest how it could be narrowed.

Prompt Template 3: “Reframe stakeholder problem statements into HMWs”

Reframe stakeholder problem statements into HMWs

Context: You’re a Product Designer translating business needs into design opportunities.
Specific Info: Stakeholders said, “We need users to complete their forms faster” and “Too many drop off on page two.”
Intent: Convert these business asks into human-centred How Might We questions.
Response Format: Show 3–4 HMW variations for each request, with explanation for how each changes the design framing.

Note where the original problem was solution-biased or not user-centred enough.

Prompt Template 4: “Validate if an HMW prompt is well-formed”

Validate if an HMW prompt is well-formed

Context: You are refining prompts from yesterday’s brainstorm.
Specific Info: You have the draft question: “How might we reduce the time it takes to update a profile?”.
Intent: Get AI feedback on clarity, openness, and design suitability of your HMW prompt.
Response Format: Score the question on key criteria and suggest 2–3 alternative phrasings with rationale.

Ask clarifying questions if goals or constraints are unclear.

Prompt Template 5: “Cluster HMWs into themes”

Cluster HMWs into themes

Context: You are preparing synthesis slides for a design sprint team kickoff.
Specific Info: You have a list of 20 HMW prompts crowdsourced from cross-functional researchers and PMs.
Intent: Group the prompts into 3–5 overarching categories or opportunity areas.
Response Format: Provide theme titles with a bullet list of aligned HMWs below each. Suggest adjacent theme overlaps, if relevant.

If any HMW seems misplaced, explain why and provide alternatives.

Prompt Template 6: “Generate HMWs from negative sentiment quotes”

Generate HMWs from negative sentiment quotes

Context: You are reviewing open-ended feedback from churned users.
Specific Info: Example quote: “I hate how many steps it takes to reset my password.”
Intent: Reframe common frustrations into progressive “How might we…” prompts to inform redesign concepts.
Response Format: Return 3 HMW questions per quote, each exploring a slightly different angle (effort, confidence, emotion).

Ask if more quote data is available if meaningful variation can’t be achieved.

Prompt Template 7: “Draft HMWs for an onboarding flow”

Draft HMWs for an onboarding flow

Context: You are redesigning the onboarding for a productivity app targeting freelancers.
Specific Info: The current flow has a 27% drop-off at account setup stage.
Intent: Develop HMW prompts that target cognitive load, form anxiety, and unclear value propositions.
Response Format: Return 5 unique How Might We questions, each tied to one UX principle.

Add a note recommending next research inputs if insights seem incomplete.

Prompt Template 8: “Turn usability findings into HMWs”

Turn usability findings into HMWs

Context: You are preparing a design iteration report.
Specific Info: Three key findings from testing: (1) Users missed key CTA buttons, (2) Tooltips were ignored, (3) Error messages felt harsh.
Intent: Translate these issues into HMW prompts to frame conceptual solutions.
Response Format: Give 2–3 HMWs per finding and include possible themes they tie into.

Ask if participant quotes are available to enrich the insight framing.

Prompt Template 9: “Create workshop-ready HMWs”

Create workshop-ready HMWs

Context: You are a facilitator designing a 60-minute ideation session with cross-functional stakeholders.
Specific Info: Chosen topic: Improve engagement after first app use.
Intent: Craft 6–8 well-balanced How Might We questions to spark creative thinking.
Response Format: Provide brief HMWs suitable for group activities, labelled for tone (e.g., playful, technical, exploratory).

Suggest ways to test these during the workshop (e.g., voting, clustering).

Prompt Template 10: “Score a batch of HMWs for impact and feasibility”

Score a batch of HMWs for impact and feasibility

Context: You are preparing prioritisation with your product manager.
Specific Info: You have 10 draft How Might We questions from a design jam.
Intent: Help sort them based on business impact and technical feasibility.
Response Format: Return a 2x2 matrix (High/Low Impact vs. High/Low Feasibility) with each HMW placed accordingly.

Clarify whether strategic priorities should influence scoring.

Recommended Tools

• Post-it notes & Pen
• Figjam or Miro for collaborative HMW workshops
•  ChatGPT or Claude for reframing and ideation support
•  Airtable or Notion to track, cluster, and prioritise HMWs collaboratively

Learn More

• NNGroup: UX Research & Design Articles
• Interaction Design Foundation – Topics
• Google Design Sprint Kit