"How Might We" questions are the pivot point of design thinking. They take the pain points and needs you discovered during empathy research and reframe them as creative challenges your team can solve. A well-written HMW question opens up a rich solution space. A poorly written one sends the team down the wrong path.

Why Three Words Matter

The phrase "How Might We" was popularized by Procter & Gamble in the 1970s and later adopted widely through IDEO and the Stanford d.school. Each word does specific work:

"How" assumes there is a solution. It shifts the team from debating whether something is possible to exploring how it could work. This subtle reframe changes the energy in a room.
"Might" gives permission to explore. It signals that we are generating possibilities, not committing to a plan. This lowers the psychological barrier to suggesting unconventional ideas. Compare "How might we..." with "How will we..." and notice how the second version already feels heavier, more final.
"We" makes it collaborative. The problem belongs to the team, not to one person. This shared ownership is especially important in cross-functional teams where no single person has the expertise to solve the problem alone.

Contrast this with other framings:

"Can we..." is binary. It invites yes/no answers rather than creative exploration.
"Should we..." implies judgment. It makes people evaluate before they generate.
"We need to..." prescribes a direction. It closes the solution space before ideation begins.
"What if we..." is close but lacks the collaborative and actionable energy of HMW.

The HMW scope spectrum: aim for questions that specify who and what, but leave room for multiple solutions

The Scope Problem

Getting the scope right is the single most important skill in writing HMW questions, and it is the most common mistake. Think of scope as a zoom level on a map:

Too Broad (Zoomed Out)

"How might we improve the user experience?" This question is so general that any idea is technically a valid answer. It provides no direction for ideation, produces scattered results, and leaves the team feeling like they brainstormed a lot but accomplished nothing.

Other examples of too-broad HMW questions:

"How might we make our product better?"
"How might we increase customer satisfaction?"
"How might we help people be more productive?"

Too Narrow (Zoomed In)

"How might we add a tooltip to the settings icon?" This is not a question; it is a solution disguised as a question. It leaves no room for creative exploration because it has already prescribed the answer. If you read a HMW question and can only think of one response, it is too narrow.

Other examples of too-narrow HMW questions:

"How might we change the button color from blue to green?"
"How might we add a search bar to the homepage?"
"How might we send a reminder email on day 3?"

Just Right (The Sweet Spot)

"How might we help new users discover key features within their first session?" This is specific enough to guide ideation (new users, first session, feature discovery) but open enough to allow dozens of different solutions (onboarding flows, interactive tutorials, progressive disclosure, gamification, contextual hints, and more).

Other well-scoped examples:

"How might we help freelancers feel confident that their invoices are tax-compliant?"
"How might we reduce the anxiety patients feel while waiting for test results?"
"How might we help remote team members build informal relationships without scheduling more meetings?"

Notice that each of these specifies who (freelancers, patients, remote team members), what (confidence, anxiety reduction, informal relationships), and an implicit constraint (without requiring tax expertise, during the waiting period, without more meetings). This structure consistently produces useful ideation prompts.

From Research to HMW: A Step-by-Step Process

Step 1: Start from Empathy Insights

Good HMW questions always come from empathy research. Pull up your empathy maps, interview notes, and observation logs. Look for:

Pain points that multiple users share
Moments of frustration, confusion, or workaround behavior
Unmet needs or desires users expressed directly or revealed through their actions
Contradictions between what users say and what they do
Emotional peaks (both positive and negative) in the user journey

Each of these is a candidate for a HMW question.

Step 2: Write the Problem Statement First

Before jumping to HMW, write a Point of View (POV) statement using the format from the Define stage:

"[User type] needs a way to [user need] because [insight from research]."

Example: "First-time project managers need a way to estimate task durations accurately because they consistently underestimate by 40 to 60 percent, leading to missed deadlines and eroded trust with their teams."

The POV statement grounds the HMW question in research. Without it, HMW questions tend to drift toward assumptions and personal opinions rather than user evidence.

Step 3: Reframe as HMW

Take your POV statement and convert it:

POV: "First-time project managers need a way to estimate task durations accurately because they consistently underestimate by 40 to 60 percent."
HMW: "How might we help new project managers create more realistic task estimates?"

Notice that the HMW drops some specificity from the POV (the exact percentage, the consequences) while keeping the core need (realistic estimates for new PMs). This is intentional. The POV captures the full context; the HMW distills it into an actionable creative prompt.

Step 4: Generate Variations

For each core insight, write 3 to 5 HMW variations that approach the problem from different angles:

Amplify the positive: "How might we make the estimation process feel like a learning opportunity rather than a guessing game?"
Remove the negative: "How might we eliminate the consequences of inaccurate estimates?"
Explore the opposite: "How might we make it acceptable for estimates to be wrong?"
Question an assumption: "How might we accomplish the project goal without fixed time estimates?"
Change the context: "How might we give PMs real-time feedback on estimate accuracy as the project progresses?"
Break the problem apart: "How might we help PMs identify which specific types of tasks they underestimate most?"

These variations are valuable because each one opens a different solution space. "Remove the negative" might lead to buffer time strategies, while "Question an assumption" might lead to no-estimate project management approaches entirely.

Step 5: Select and Prioritize

You will typically generate 15 to 30 HMW questions from a research synthesis session. You cannot ideate on all of them. Select 3 to 5 for your ideation session based on:

User impact: If solved, how much would this improve the user's experience?
Business alignment: Does this connect to goals and metrics the organization cares about?
Feasibility window: Is a meaningful solution possible given current constraints?
Team energy: Which questions make the team lean forward? Genuine curiosity produces better ideation than obligatory brainstorming.

Dot voting works well for this. Give each team member 3 votes. The questions with the most votes become your ideation prompts.

Domain-Specific Examples

E-commerce

"How might we help shoppers feel confident about fit without trying items on?"
"How might we reduce the anxiety of buying expensive items from unfamiliar brands online?"
"How might we make product returns feel like a positive experience rather than a hassle?"

Healthcare

"How might we help patients remember their medication routine without feeling nagged?"
"How might we make medical information understandable without oversimplifying the science?"
"How might we help caregivers coordinate care across multiple specialists without becoming full-time project managers?"

Education

"How might we help students recognize what they do not understand before the exam?"
"How might we make homework feel purposeful rather than punitive?"
"How might we help teachers identify struggling students before they fall too far behind?"

B2B Software

"How might we help new employees become productive with our tool within their first week?"
"How might we reduce the time managers spend generating reports without sacrificing data quality?"
"How might we make cross-team collaboration feel natural rather than like an overhead task?"

Using HMW Questions in Ideation

Once you have selected your top HMW questions, use each one as a prompt for a focused brainstorming session. Write the HMW question where everyone can see it. Set a timer for 8 to 10 minutes. Have each person generate ideas silently on sticky notes or in a shared document (silent brainstorming prevents groupthink). Then share, discuss, and build on each other's ideas.

The specific framing of each HMW question naturally guides ideation toward relevant solutions. A well-scoped HMW question makes brainstorming feel productive rather than aimless, which is often the difference between a team that generates useful ideas and one that spins in circles. See the Ideate stage guide for detailed brainstorming techniques.

AI-Assisted HMW Generation

AI tools like Design Thinker Labs can generate HMW questions from your empathy research data and project context. The AI analyzes your empathy maps, interview notes, and problem statements, then produces a set of HMW questions at varying scopes. This gives your team a strong starting point for discussion and refinement, especially when you are new to the technique and want to see what well-crafted HMW questions look like for your specific problem.