Discovery Process (End-to-End)
Run a complete 6-phase discovery cycle — frame, research, synthesize, generate, validate, decide — in 3-4 weeks. Orchestrates 10+ component skills for problem-to-validated-solution workflow.
You keep running "discovery" that's really just scheduling a few customer calls and declaring victory. A real discovery cycle has six phases, three decision points, and ends with GO / PIVOT / KILL — not a vague Confluence doc nobody reads. 3–4 weeks, not 3–4 months.
Who it's for: PMs running continuous discovery, product leads preventing "build it and they will come" mistakes, Heads of Product investigating churn, founders validating before building, cross-functional teams aligning before roadmap commitment
Example
"Our activation rate dropped 15% this quarter and I need to run discovery to find out why" → Phase 1 framing + hypothesis + Phase 2 interview plan (5-10 customers) + Phase 3 conduct + Phase 4 affinity map + top 3 pain points + Phase 5 opportunity solution tree + 2 experiments + Phase 6 GO/PIVOT/KILL decision + epics/PRD
CLAUDE.md Template
New here? 3-minute setup guide → | Already set up? Copy the template below.
npx ccplaybooks add discovery-process# Discovery Process
Guide product managers through a complete discovery cycle—from initial problem hypothesis to validated solution—by orchestrating problem framing, customer interviews, synthesis, and experimentation into a structured 6-phase process.
This is not a one-time research project. It's a continuous discovery practice that runs in parallel with delivery, typically 1–2 cycles per quarter.
## The Six Phases (Teresa Torres, Marty Cagan)
1. **Frame the Problem** — Define what you're investigating and why
2. **Conduct Research** — Gather qualitative and quantitative evidence
3. **Synthesize Insights** — Identify patterns, pain points, opportunities
4. **Generate Solutions** — Explore multiple optionsGet new playbooks like this one
One email a week with new Claude Code workflows. Free, like everything here.
No spam. Unsubscribe anytime.
What This Does
Orchestrates a complete 6-phase discovery cycle — Frame → Research → Synthesize → Generate → Validate → Decide — over 3-4 weeks. Pulls together problem framing, interviews, synthesis, solution trees, and experiments into a structured process with 3 decision points (enough context? saturated? validated?).
Not a one-time research project. A repeatable cycle designed to run 1-2 times per quarter alongside delivery.
Quick Start
mkdir -p ~/Documents/DiscoveryCycle
mv ~/Downloads/CLAUDE.md ~/Documents/DiscoveryCycle/
cd ~/Documents/DiscoveryCycle
claude
Describe your hypothesis or problem area. Claude guides you phase-by-phase, referencing other skills (problem-framing, discovery-interview-prep, opportunity-solution-tree, epic-hypothesis, prd-development) as it goes.
The Six Phases
| Phase | Duration | Output |
|---|---|---|
| 1. Frame the Problem | 1–2 days | Problem hypothesis + research questions + success criteria |
| 2. Research Planning | 1 day | Interview guide + 5–10 recruited participants |
| 3. Conduct Research | 1–2 weeks | 5–10 interviews + support/analytics themes |
| 4. Synthesize Insights | 2 days | Affinity map + top 3–5 prioritized pain points |
| 5. Generate & Validate | 1–2 weeks | 3–9 solutions + experiment results |
| 6. Decide & Document | 2–3 days | GO/PIVOT/KILL + epics + PRD + exec alignment |
Three Decision Points
- After framing: Enough context to start research, or gather existing data first?
- After interviews: Reached saturation (same patterns across 3+), or keep interviewing?
- After experiments: Solution validated, pivot, or kill?
Orchestrated Skills
problem-framing-canvas,problem-statement,proto-persona,jobs-to-be-done(Phase 1)discovery-interview-prep(Phase 2)customer-journey-mapping-workshop(Phase 4, optional)opportunity-solution-tree,lean-ux-canvas(Phase 5)epic-hypothesis,prd-development(Phase 6)
Tips & Best Practices
- Time-box to 3–4 weeks. Longer = analysis paralysis; ship insights, then iterate.
- Interview 5–10, not 2–3. Saturation matters; small samples = confirmation bias.
- Past behavior, not wishes. Mom Test discipline throughout Phase 3.
- Run continuous discovery between cycles. One interview per week, ongoing — don't only discover when a big initiative starts.
- Always end with GO/PIVOT/KILL. A cycle that ends with "let's think about it" wasted 3 weeks.
Common Pitfalls
- Skipping customer interviews — analytics alone misses the "why"
- Asking leading questions that confirm your hypothesis
- Stopping at 1–2 interviews before saturation
- Spending 6 weeks synthesizing instead of moving to experiments
- Running discovery once before a launch, then stopping — it's continuous, not one-time