Problem Framing Canvas — Identify the Right Problem Before Developing a Solution

Skill Overview

The Problem Framing Canvas is an interactive AI skill based on the MITRE Innovation Toolkit. Through structured questioning across three stages—“Look Inward,” “Look Outward,” and “Reframe”—it helps product managers confirm that they are solving the right problem before investing in a solution. The final output is a clear problem statement and a How Might We (HMW) question.

Use Cases

1. The problem is vague and needs to be clarified: For example, “mobile retention is declining, but we can’t explain why.” The skill starts with symptoms, stakeholders, and existing research, breaking down vague dissatisfaction into an analyzable structure.
2. Stakeholders have already proposed a solution: When a leader or business stakeholder makes a solution-oriented request such as “add gamification to the CRM,” the skill makes the team’s solution fixation explicit and works backward to uncover the real problem behind it, preventing the team from accepting the request without question.
3. The team needs alignment on the problem definition: When cross-functional team members have different perspectives and are preparing to begin discovery research for a new project, the skill can guide the team to an anti-bias problem statement within 30–45 minutes, providing a shared starting point for subsequent brainstorming and solution exploration.

Core Functions

1. Three-stage structured guidance: The skill follows the three stages of the MITRE Canvas—“Look Inward” to examine one’s own assumptions and biases, “Look Outward” to understand who experiences the problem, who is excluded, and who benefits from the status quo, and “Reframe” to synthesize and produce a new formulation. The process proceeds through one question at a time in a gradual, step-by-step manner.
2. Examination of assumptions and biases: It specifically asks, “Are we part of the problem?” to identify confirmation bias (assuming we understand users without validating it), survivorship bias (focusing only on mainstream users), and solution-first thinking (jumping to “we need to add a feature” without clarifying the root cause). It also highlights whether the problem has remained unsolved because it is difficult, considered unimportant, or threatens someone’s interests.
3. Output of a problem statement and HMW question: The final output is a structured canvas document containing a complete restatement of “who experiences what problem, in what context, with what consequences, and why it has continued to be overlooked,” along with a well-scoped “How Might We…” question that can directly lead into next steps such as an opportunity-solution tree or user interviews.

Frequently Asked Questions

How is it different from brainstorming or solution discussions?

The Problem Framing Canvas is solely responsible for defining the problem; it does not generate solutions. It is intentionally used before solution discussions to prevent the team from entering with a predetermined solution. If the problem has already been validated and only ideas are needed, proceed directly to brainstorming or an opportunity-solution tree instead of using this canvas.

What should I prepare before using it? Can I use it alone?

You can start from scratch—the skill begins by asking about your initial understanding of the problem. If available, providing a problem description, retention or support-ticket data, existing research, and the “solution everyone has already agreed on” will improve the results. The skill will skip anything that has already been answered. It supports individual use, but it is recommended that you review the completed canvas with a cross-functional team to minimize lingering personal bias.

When is this Problem Framing Canvas not appropriate?

There are three situations: the problem has already been thoroughly validated and contains no ambiguity; you are simply fixing a clearly defined bug or addressing technical debt, with no need for deep reframing; or stakeholders have already finalized the solution. In the latter case, forcing the canvas is likely to become a formality, so it is better to first align on the solution itself.

What should I do with the output afterward?

The HMW question generated by the canvas can be used directly for solution brainstorming. The restated problem statement can be passed to a problem-statement skill to be developed into the problem definition for a PRD or roadmap. Alternatively, you can first validate through user interviews whether the reframed understanding holds true.