How Might We
A reframing technique popularised by IDEO and the Stanford d.school that transforms problem statements and observations into open-ended ideation prompts. Each 'How Might We...' question is specific enough to provide direction but open enough to allow creative responses. HMW questions bridge the gap between problem definition and ideation in Design Thinking.
Facilitation script
- 1
Post the insights or problem statements from prior research on the wall and walk through them so everyone shares the same evidence base.
5 min - 2
Teach the formula — 'How might we [verb] [subject] so that [outcome]?' — and demo one live reframing: 'Users abandon onboarding at step 3' becomes 'How might we show users the value before step 3?'
5 min - 3
Have everyone silently write several HMW questions per insight, one per sticky note. Push for quantity — the first framing is rarely the best one.
10 min - 4
Post and cluster the questions on the wall, grouping duplicates and near-duplicates.
5 min - 5
Quality-check the clusters together: flag questions that are too broad to focus energy or so narrow they smuggle in a solution, and rewrite them as a group.
5 min - 6
Dot vote on the most promising questions and carry the top two or three into your brainstorming or ideation session.
5 min
Tips
HMW questions can be too broad ('How might we improve the product?') or too narrow ('How might we add a progress bar to step 3?').
The sweet spot is specific enough to focus energy but broad enough to allow surprising answers.
Common pitfalls
Writing HMW questions so broad ('How might we improve the product?') that the following brainstorm sprays in every direction and produces nothing buildable
Embedding the solution in the question ('How might we add a progress bar to step 3?') — the brainstorm collapses into speccing one predetermined feature
Allowing only one HMW per insight — the group anchors on the first framing, which is rarely the strongest; insist on several per insight
Skipping the vote and carrying every question into ideation, which spreads a 45-minute session too thin to develop any answer properly
The facilitator wordsmithing the group's questions into their own phrasing — participants stop owning the challenges they're about to ideate against
Variations
Use HMW as a standalone workshop to generate a bank of design challenges. Combine with Crazy 8s for rapid ideation against the top questions.
Where it fits
When to use it
You're bridging from research to ideation — interviews or observation have produced insights that need converting into brainstorm-ready prompts
The team keeps jumping from a problem statement straight to a single pet solution and the frame needs opening back up
A fuzzy challenge needs breaking into several concrete angles of attack before a Crazy 8s or brainstorming round
Stakeholders disagree about what the problem actually is — writing and clustering HMW questions makes the competing framings visible
You want a standalone session that builds a reusable bank of design challenges for future sprints
When not to use it
No research or problem definition exists yet — HMW questions reframe insights, they can't substitute for them; run interviews or observation first
The solution is already decided and the session would only be reframing theatre — be honest and plan implementation instead
The real need is to test what the team believes rather than generate ideas — Assumption Storming attacks the underlying beliefs directly
The question is a closed technical decision with a discoverable right answer — analysis beats open ideation prompts there
Related methods
Further reading
Frequently asked questions
What is a How Might We question?▾
A How Might We (HMW) question is a reframing device from Design Thinking, popularised by IDEO and the Stanford d.school, that turns a problem statement or research insight into an open-ended ideation prompt. A good HMW is specific enough to give a brainstorm direction but open enough to allow surprising answers, bridging the gap between problem definition and ideation.
How do you write a good How Might We question?▾
Use the formula 'How might we [verb] [subject] so that [outcome]?' and write several candidates per insight rather than settling on the first. Then test each against the sweet spot: too broad ('How might we improve the product?') gives no direction, while too narrow ('How might we add a progress bar to step 3?') smuggles in a solution and kills creative range.
What is an example of a How Might We reframing?▾
Take the observed problem 'Users abandon onboarding at step 3.' Two valid HMW reframings are 'How might we make step 3 feel effortless?' and 'How might we show users the value before step 3?' — the same insight opens two different solution spaces, which is why you write several questions per insight and vote on the best.
Why say 'how might we' instead of 'how can we'?▾
'Might' deliberately lowers the stakes: it signals that answers are candidates to explore, not commitments, which frees people to suggest ideas that could fail. 'How' presumes solutions exist and 'we' makes the challenge collaborative — together the phrasing keeps ideation open where 'how can we' or 'how should we' invite feasibility debates too early.
How long does a How Might We session take?▾
Plan 20–45 minutes for the core sequence: presenting insights, individually writing questions, clustering, and dot voting on the top prompts. It works with anywhere from 2 to 30 people and needs only sticky notes and markers; the winning questions then feed a separate ideation round such as Crazy 8s.
Plan your next workshop with AI
Workshop Weaver helps you combine methods like How Might We into a complete, timed agenda in minutes.
Try it freeMethod descriptions on Workshop Weaver are original content written by our team, based on established facilitation practices.