How Might We

Overview

The way you phrase a problem shapes what solutions feel possible. "How Might We" is a three-word construction that sounds almost too simple to matter. It reframes a challenge as an invitation rather than a directive, keeping the problem space wide enough for creative thinking without dissolving into vagueness.

Each word is doing something. "How" assumes a solution exists. "Might" holds it loosely, implying that several approaches are worth exploring. "We" makes it shared. Together they create a question that's specific enough to act on and open enough to surprise you.

The format traces back to Min Basadur at Procter & Gamble in the 1970s and was later adopted by IDEO, where it became a standard bridge between research and ideation. Its staying power comes from something straightforward: questions generate different thinking than statements do. A problem statement tells your team what's wrong. An HMW question asks what could be possible.

The trap is treating it as a reformatting task. The point isn't to convert your problem statement into question form. It's to interrogate what you actually want to solve by generating many versions of the question, then choosing the framing that opens the most productive space.

When to Use It

• After user research or affinity mapping, when you've identified themes and want to convert them into prompts for ideation.
• At the start of a workshop or sprint to establish a shared problem frame before generating ideas.
• When your team is converging on a solution too quickly and needs to step back and examine what problem you're actually solving.
• When a brief has arrived from leadership with a solution already baked into the problem statement and you need to open it back up.

Skip it when you already have strong alignment on the problem and a clear direction. Adding an HMW step to a well-defined project can feel like backtracking. It earns its place when the problem space is still fuzzy.

How It Works

Start from something real: a user quote, a friction point from research, a recurring theme from an affinity session. HMW questions generated from instinct are weaker than ones grounded in evidence.

Write the question beginning with "How might we..." and complete it with a verb phrase that describes the change you're after. One idea per question. Keep an eye on altitude. If the question is so broad it could mean almost anything ("How might we improve the onboarding experience?"), zoom in. If the answer is already implied ("How might we add a progress indicator to step three?"), zoom out. Aim for the space between: specific enough to generate concrete ideas, open enough that more than one solution could answer it.

Generate multiple versions before committing to one. A strong prompt set for an ideation session might be six to ten HMW questions drawn from the same cluster of insights. The variety gives ideation more to work with and helps the team notice which framings feel most generative.

Tips

Don't write one and move on. Teams that take time to generate and compare multiple versions consistently land on better framing. The act of choosing is part of the thinking.

Watch for solution creep. If your question already implies the answer, you've skipped ideation without realizing it. "How might we send a confirmation email after checkout?" is a feature request in question clothing, not an HMW.

Source your questions from real evidence. The best ones trace directly to something a user said or a pattern you observed. When you can point to what generated a question, the ideation that follows feels grounded, not arbitrary.

Resist safe framing. Aspirational questions ("How might we make users feel like experts within their first week?") tend to generate more interesting ideas than cautious ones.

The Output

A set of problem frames, usually five to ten per session, that your team can take directly into ideation. Each question acts as a prompt: open-ended, grounded, and free of embedded answers.

HMW questions typically feed into brainstorming or concept sketching. They also work well as anchors in a design brief, giving the team something to return to when conversations drift toward scope creep.

Related Methods

• Affinity Mapping: Comes before. Themed clusters from an affinity session are natural raw material for HMW questions; one cluster can easily generate several.
• Interviewing: Comes before. Direct quotes and observed pain points make the best source material for grounded HMW questions.
• Rose, Thorn, Bud: Comes before. Thorns and buds both point to unmet needs worth reframing as opportunities.
• Usability Testing: Can come before. Pain points surfaced in testing translate naturally into HMW prompts for the next design iteration.
• Target Prioritization: Comes after. Once ideation produces options, prioritization helps the team decide which ideas to pursue.
• Impact/Effort Matrix: Comes after. Pairs with Target Prioritization to evaluate which ideas are worth moving forward.