Product framing

Product Framing

Use before interface recommendations when problem, user outcome, segment, or evidence is unclear. Goal: prevent designing the wrong screen for the wrong reason, not create workshop artifacts.

Core Model

Frame:

1. User/context: actor, situation, constraints.
2. Need/job: progress sought.
3. Outcome: observable user or business state to improve.
4. Opportunity: pain, motivation, barrier, or unmet need.
5. Solution bet: candidate interface or behavior change.
6. Evidence: research, analytics, support, observation, or source.
7. Metric/counter-metric: value signal and guardrail.

Evidence strength: user behavior, usability sessions, analytics, support logs, and production traces outrank stakeholder preference; primary standards/platform guidance outrank blog summaries; heuristics are risk signals, not proof.

Frameworks:

• Double Diamond: discover evidence, define problem/success, develop bets, deliver smallest testable intervention.
• Opportunity Solution Tree: outcome -> opportunities -> solution bets -> evidence/criteria.
• HEART/GSM: happiness, engagement, adoption, retention, task success; goal -> signal -> metric; counter-metric.

Questions To Resolve

Ask only if missing context would materially change the recommendation:

• Which user segment is this for?
• What task or outcome matters most?
• Is the main risk desirability, usability, feasibility, viability, or trust?
• What evidence already exists?
• What existing product or workflow should this preserve?

Output

Produce:

1. Problem frame: one compact problem statement.
2. Primary user outcome: observable and specific.
3. Assumptions: ranked by risk.
4. Opportunity map: 3-7 opportunities, not features.
5. Solution principles: constraints for interface work.
6. Metric plan: goal, signal, metric, counter-metric.
7. Smallest next step: design, research, prototype, or measurement action.

For durable product decisions, use design_intelligence.decision.v1 from references/contracts.md.

Quality Bar

Good framing makes weak UI asks sharper. If it still sounds like "make it cleaner" or "improve UX," continue framing.

Avoid:

• generic personas without behavior evidence;
• invented metrics;
• feature lists disguised as opportunities;
• business-only goals with no user outcome;
• user-only goals with no product constraint;
• engagement metrics without counter-metrics.