Discovery Process

Purpose

Run a complete discovery cycle from problem framing through validated solutions. This workflow orchestrates multiple skills into a structured process that ensures you understand the problem, validate it with real users, and only commit to solutions backed by evidence.

Key Concepts

Discovery Process Overview

Week 1: UNDERSTAND
├─ Phase 1: Frame the Problem (Day 1-2)
├─ Phase 2: Plan Research (Day 3)
└─ Decision Point: Ready to research?

Week 2: RESEARCH
├─ Phase 3: Conduct Research (Week 1-2)
└─ Decision Point: Reached saturation?

Week 3: SYNTHESIZE & VALIDATE
├─ Phase 4: Synthesize Insights (End of Week 2)
├─ Phase 5: Generate & Validate Solutions (Week 3)
└─ Decision Point: Solution validated?

Week 4: DECIDE
└─ Phase 6: Decide & Document (End of Week 3-4)

Anti-Patterns

• Not a one-time event: Discovery is continuous, not a "project phase"
• Not optional: Discovery isn't a luxury — it's the cheapest way to avoid building the wrong thing
• Not just interviews: Combines qualitative (interviews) + quantitative (analytics, surveys) evidence

Application

Phase 1: Frame the Problem (Day 1-2)

Goal: Define what problem you're investigating and for whom.

Activities:

Outputs:

• Initial problem statement (hypothesis)
• 2-3 proto-personas
• JTBD statements
• Reframed problem (if applicable)

Decision Point 1: Do we have enough context to start research?

• ✅ YES → Proceed to Phase 2
• ❌ NO → Spend 1 more day on desk research and stakeholder interviews

Phase 2: Plan Research (Day 3)

Goal: Design the research plan — who to talk to, what to ask, how to measure.

Activities:

Outputs:

• Interview guide (5-8 interviews planned)
• Survey draft (if applicable)
• Competitive landscape overview

Phase 3: Conduct Research (Week 1-2)

Goal: Talk to real users and gather evidence.

Activities:

• Conduct 5-8 discovery interviews
• Analyze behavioral analytics (funnels, drop-offs)
• Review support tickets and NPS verbatims
• Run validation survey (if sample size allows)

Outputs:

• Interview transcripts/notes with key quotes
• Analytics insights
• Survey results (if applicable)

Decision Point 2: Have we reached saturation?

• ✅ YES (hearing same themes from multiple sources) → Proceed to Phase 4
• ❌ NO → Schedule 3-5 more interviews; expand to adjacent segments

Phase 4: Synthesize Insights (End of Week 2)

Goal: Turn raw data into actionable themes.

Activities:

• Cluster interview insights into themes (affinity diagram)
• Update proto-personas with validated attributes
• Map the customer journey based on actual data
• Identify the top 3-5 validated pain points

Outputs:

• Insight themes (clustered, prioritized)
• Updated proto-personas
• Validated customer journey map
• Refined problem statement (incorporating evidence)

Phase 5: Generate & Validate Solutions (Week 3)

Goal: Generate solution hypotheses and test them lightly.

Activities:

Outputs:

• Opportunity-solution tree
• 2-3 solution hypotheses with success criteria
• Lightweight validation results (concept tests, prototypes, or fake doors)

Decision Point 3: Did experiments validate a solution direction?

• ✅ YES → Proceed to Phase 6
• ❌ NO → Iterate on solution hypotheses or revisit problem framing

Phase 6: Decide & Document (End of Week 3-4)

Goal: Commit to a direction and package findings for stakeholders.

Activities:

• Write the final discovery report
• Update the problem statement to "validated" status
• Create the initial opportunity backlog
• Present findings to stakeholders

Outputs:

• Discovery report (findings, recommendations, evidence)
• Validated problem statement
• Opportunity backlog (prioritized)
• Stakeholder presentation deck

Complete Workflow Summary

Week 1:
├─ Day 1-2: Frame the Problem
│  ├─ problem-statement (30 min)
│  ├─ proto-persona (30-60 min)
│  ├─ jobs-to-be-done (30 min)
│  └─ problem-framing-canvas (45 min)
│
├─ Day 3: Plan Research
│  ├─ discovery-interview-prep (30 min)
│  ├─ survey-design (30 min)
│  └─ company-research (45 min)
│
└─ Decision: Ready to research?

Week 2:
├─ Conduct 5-8 interviews
├─ Analyze behavioral data
├─ Run validation survey
└─ Decision: Reached saturation?

Week 3:
├─ Synthesize insights
│  ├─ customer-journey-map (45 min)
│  └─ Cluster themes + update personas
│
├─ Generate & validate solutions
│  ├─ opportunity-solution-tree (60 min)
│  └─ epic-hypothesis (30 min each)
│
└─ Decision: Solution validated?

Week 4:
└─ Document & present
   ├─ Final discovery report
   ├─ Opportunity backlog
   └─ Stakeholder presentation

Total estimated effort: 25-40 hours over 3-4 weeks

Common Pitfalls

Pitfall 1: Skipping Customer Interviews

Symptom: "We already know what users want from support tickets."

Fix: Support tickets are biased toward vocal users with acute problems. Interviews reveal quiet majority needs.

Pitfall 2: Not Reaching Saturation

Symptom: After 3 interviews, "We've heard enough."

Fix: Saturation typically requires 5-8 interviews per segment. If themes are still emerging, keep going.

Pitfall 3: Discovery Without Decision Points

Symptom: Teams run discovery forever, never transitioning to execution.

Fix: Each phase has an explicit decision point. If the answer is "yes," move forward. Set timeboxes.

Pitfall 4: Treating Discovery as a One-Time Activity

Symptom: "We did discovery last quarter, we don't need to do it again."

Fix: Discovery is continuous. Markets change, users evolve, assumptions expire. Run mini-discovery cycles quarterly.

References

Skills Orchestrated

• skills/discovery/problem-statement/SKILL.md
• skills/discovery/proto-persona/SKILL.md
• skills/discovery/jobs-to-be-done/SKILL.md
• skills/discovery/problem-framing-canvas/SKILL.md
• skills/discovery/discovery-interview-prep/SKILL.md
• skills/discovery/survey-design/SKILL.md
• skills/discovery/company-research/SKILL.md
• skills/discovery/customer-journey-map/SKILL.md
• skills/discovery/customer-journey-mapping-workshop/SKILL.md
• skills/specification/opportunity-solution-tree/SKILL.md
• skills/planning/epic-hypothesis/SKILL.md

External Frameworks

• Teresa Torres, Continuous Discovery Habits (2021)
• Rob Fitzpatrick, The Mom Test (2013)
• Marty Cagan, Inspired (2017)

Skill type: Workflow Domain: Discovery & Research