How might we questions
Turn a need you have evidence for into one open question your team can answer in many ways.
- Time
- 20 min
- People
- 3–6
- Kind of work
- Frame · Generate
In short
A how might we question rewrites a real, evidenced need as an open challenge, such as "How might we help first-time applicants know which documents to bring?" To run it, pick one need from your research, name the person and the constraint, draft several questions, then strip out any solution hiding in the wording.
What it is
How might we is a short question format that sits between research and ideas. You take something you learned about people, such as a need, a frustration or a workaround, and phrase it as a challenge that starts with "How might we". The three words do quiet work. "How" assumes an answer exists. "Might" gives permission to suggest things that may not work. "We" makes it a shared job.
A good question is narrow enough to point at one person and one need, and wide enough to allow ten different answers. The format is widely used in design practice and is well known through human-centred design teams. In this worksheet you record four things: the person, the need, the constraint you must respect, and the final opportunity question.
When to use it
Use it when
- You have finished interviews or observation and need to agree what problem to work on next.
- A team keeps jumping to one favourite solution and you want to reopen the options.
- You are about to run an idea session and need a clear prompt to put on the wall.
- A brief from leadership is vague, such as "improve the customer experience", and you need to make it workable.
- You want to split a large challenge into several smaller ones that different groups can take.
Skip it when
- You have no evidence yet about the people involved. Do some research first, or the question will rest on guesses.
- The answer is already decided and approved. Writing open questions then only creates false hope.
- The issue is a simple fix with one obvious owner, such as a broken link on a web form.
You need
- Time
- About 20 minutes for one need, longer if you draft questions for several needs.
- People
- 3 to 6 people, ideally including someone who did the research.
- Materials
- Research notes or a point of view statement, sticky notes or a shared board, and markers.
How to run it
Step 1: Choose one evidenced need
Put your research summary on the wall and pick a single need that came up more than once. Write it in plain words on the worksheet next to the person who has it, for example "a first-time applicant who does not know which documents count". If the team cannot point to a quote or an observation behind it, choose another need.
Step 2: Name the constraint
Write down the one limit any answer must respect. It might be a legal rule, a budget, a deadline or a fixed system. Naming it now stops the group from producing ideas that cannot be used, and it keeps the question honest.
Step 3: Draft questions alone
Give everyone five minutes to write three to five questions on separate notes, each starting with "How might we". Ask for variety: one that is broad, one that is narrow, one that flips the need, one that focuses on a moment in the journey. Working alone first means quieter people get their wording on the wall.
Step 4: Read them out and group them
Each person reads their questions without explaining them. Put similar ones together. You will usually end up with three or four families of questions that point in different directions.
Step 5: Remove hidden solutions
Check each strong question for a solution tucked inside. "How might we build an app that reminds applicants" already assumes an app. Rewrite it as "How might we help applicants remember what to bring". If the question only has one possible answer, it has become an instruction.
Step 6: Test the width
Take your best two questions and spend two minutes listing quick answers for each. If you get fewer than five, the question is too narrow. If the answers are wildly unrelated to the need, it is too broad. Adjust the wording until it sits in the middle.
Step 7: Agree one question and record it
Choose the question the team will take into idea generation and write it in the opportunity question field. Keep the runners-up on the worksheet. They are useful for a second round or another team.
Worked example
A municipal licensing department found that about a third of shop licence applications were returned because a document was missing. Interviews showed that first-time applicants, often small family businesses, did not know which version of a lease or ID copy was accepted.
Five people met for 20 minutes: two counter staff, a team lead, a digital officer and the researcher who ran the interviews. They chose the need "know before the visit which documents will be accepted".
The constraint was that the required document list is set by regulation and cannot change this year.
Silent drafting produced 17 questions. Two of them hid solutions, such as "How might we add a checklist to the website". The group rewrote that one as "How might we help applicants check their documents before they leave home".
A quick width test gave nine possible answers for that question, from a phone call the day before to a sample document pack at partner offices. The team kept it as their main question and stored two others for a later session.
IllustrativeTips and mistakes
Tips
- Keep the person in the question. "How might we help a first-time applicant" beats "How might we improve applications".
- Write several and choose. The first question anyone writes is rarely the best one.
- Use the words people used in interviews, so the question still sounds like them.
- Pin the final question where everyone can see it during the idea session that follows.
Common mistakes
- Writing the question before agreeing the need, so the team argues about wording when it really disagrees about the problem.
- Making it so broad, like "How might we make government better", that nobody knows where to start.
- Slipping a favourite solution into the wording and calling it a question.
- Dropping the constraint, which leads to exciting ideas that cannot be delivered.
What you walk away with
You leave with one agreed opportunity question, tied to a named person, a need you have evidence for and a constraint everyone accepts, plus two or three backup questions. The next step is to generate ideas against it. Crazy eights works well for quick visual sketching, and brainwriting suits groups where a few voices tend to dominate.
Questions people ask
How long does a how might we session take?
About 20 minutes for one need with a small group. If you want questions for four or five needs, allow 45 to 60 minutes and take a short break in the middle.
What is the difference between a how might we question and a problem statement?
A problem statement describes what is wrong and for whom. A how might we question takes that statement and turns it into an invitation to find answers. Many teams write a point of view statement first and then turn it into one or more questions.
How many how might we questions should we end with?
Choose one to work on now. Keep two or three alternatives on record so you can return to them if the first round of ideas runs dry.
Tool card
The worksheet your team fills in, one box per field.
- Person
- Need
- Constraint
- Opportunity question