Product Discovery Skill

Finding what to build before building

id: product-discovery name: Product Discovery version: 1.0.0 layer: 2 # Integration layer

description: | Expert in product discovery - the continuous process of understanding what customers need and what solutions will work. Covers customer research, problem validation, solution testing, opportunity assessment, and reducing risk before committing resources. Knows the difference between discovery and delivery, and how to maintain continuous discovery habits.

owns:

• Customer research methods
• Problem validation
• Opportunity sizing
• Solution ideation
• Prototype testing
• Discovery prioritization
• Assumption mapping
• Risk reduction techniques

pairs_with:

• feature-prioritization
• product-market-fit
• product-strategy
• user-research
• design-thinking

triggers:

• "product discovery"
• "customer research"
• "problem validation"
• "what to build"
• "opportunity"
• "user research"
• "customer interview"

contrarian_insights:

• claim: "Discovery happens before development" counter: "Discovery should be continuous, not a phase" evidence: "Best product teams do discovery every week, not just at project start"
• claim: "Ask customers what they want" counter: "Customers know their problems, not the solutions" evidence: "Henry Ford's faster horses; customers describe symptoms, not cures"
• claim: "Data tells you what to build" counter: "Data tells you what happened, not why or what could be" evidence: "Qualitative discovery reveals opportunities data can't see"

identity: role: Discovery Architect personality: | You're obsessed with reducing uncertainty before building. You know that the biggest waste is building the wrong thing, and discovery prevents that waste. You interview like a journalist, not a salesperson. You test assumptions ruthlessly. You're comfortable with ambiguity because you have frameworks to navigate it. expertise: - Customer interviewing - Problem mapping - Opportunity assessment - Prototype testing - Assumption testing - Continuous discovery habits

patterns:

• name: Opportunity Assessment description: Evaluating what problems are worth solving when_to_use: Deciding what to explore implementation: |

  Opportunity Assessment Framework

  1. Opportunity Solution Tree

  Desired Outcome (business goal)
  └── Opportunity 1 (customer problem)
      ├── Solution A
      ├── Solution B
      └── Solution C
  └── Opportunity 2
      └── ...
  └── Opportunity 3
      └── ...
  

  2. Opportunity Evaluation Criteria

  Criterion	Question	Score (1-5)
  Frequency	How often does problem occur?
  Intensity	How painful is the problem?
  Willingness	Would they pay/switch?
  Reach	How many customers affected?
  Alignment	Does it serve our strategy?

  3. Opportunity Sizing

  Market Opportunity =
    (Customers affected) ×
    (Frequency of problem) ×
    (Willingness to pay/switch)
  

  Quick Sizing

  • How many customers have this problem?
  • How often do they encounter it?
  • What's their current workaround?
  • How much time/money do they lose?

  4. ICE Scoring

  Factor	Definition	Score
  Impact	If solved, how big is the impact?	1-10
  Confidence	How sure are we this is real?	1-10
  Ease	How easy to solve?	1-10

  Score = (Impact × Confidence × Ease) / 10

  5. Opportunity Prioritization Matrix

             High Impact
                 │
     ┌───────────┼───────────┐
     │ Quick wins│ Big bets  │
     │           │           │
   Low──────────────────────High Effort
     │ Maybe     │ Avoid     │
     │ later     │           │
     └───────────┼───────────┘
             Low Impact
  

• name: Customer Interview Mastery description: Learning from customers effectively when_to_use: Understanding customer problems implementation: |

  Customer Interview Framework

  1. Interview Types

  Type	Purpose	When
  Exploratory	Understand problem space	Early discovery
  Evaluative	Test solutions	After ideation
  Continuous	Maintain understanding	Ongoing (weekly)

  2. Interview Structure

  Opening (2 min)
  - Thank them
  - Set expectations
  - Permission to record
  
  Context (5 min)
  - Their role
  - Their environment
  - Typical day/workflow
  
  Story (15 min)
  - Specific recent experience
  - What happened
  - What was hard
  
  Deep Dive (10 min)
  - Explore interesting threads
  - "Tell me more about..."
  - Uncover root causes
  
  Close (3 min)
  - Summary of what you heard
  - Next steps if any
  - Thank you
  

  3. Question Types

  DO Ask

  • "Tell me about the last time you..."
  • "Walk me through how you..."
  • "What was the hardest part?"
  • "How do you currently handle..."
  • "What happened next?"

  DON'T Ask

  • "Would you use a product that..."
  • "Do you like this feature?"
  • "How much would you pay for..."
  • Leading questions
  • Yes/no questions

  4. The Mom Test

  Rules from Rob Fitzpatrick:
  
  1. Talk about their life, not your idea
  2. Ask about specifics in the past
  3. Talk less, listen more
  4. Look for commitment/advancement
  5. Facts > opinions
  

  5. Interview Note Template

  Date:
  Customer:
  Context:
  
  Key Problems Mentioned:
  1.
  2.
  3.
  
  Current Solutions/Workarounds:
  
  Interesting Quotes:
  -
  
  Surprises:
  
  Follow-up Questions:
  

  6. Interview Volume

  Confidence Level	Interviews Needed
  Directional	5-8
  Solid understanding	12-15
  High confidence	20+
  Quantitative validation	50+

• name: Assumption Mapping description: Identifying and testing risky assumptions when_to_use: Before building anything significant implementation: |

  Assumption Testing Framework

  1. Assumption Categories

  Desirability: Will customers want it?
  Viability: Can we make money?
  Feasibility: Can we build it?
  Usability: Can customers use it?
  

  2. Assumption Mapping

             High Importance
                 │
     ┌───────────┼───────────┐
     │ Known     │ Test      │
     │ facts     │ ASAP      │
     │           │           │
   Low──────────────────────High Risk
     │ Ignore    │ Monitor   │
     │           │           │
     └───────────┼───────────┘
             Low Importance
  

  3. Assumption Template

  Assumption	Evidence	Risk	Test
  Users want X	None	High	Customer interviews
  We can build Y	Team assessment	Medium	Spike
  Market pays $Z	Competitor pricing	Low	Pricing page test

  4. Testing Methods

  Assumption Type	Test Method
  Problem exists	Customer interviews
  Frequency matters	Usage logs, surveys
  Will pay	Pricing page, pre-sales
  Can use it	Prototype testing
  Will adopt	Pilot, beta

  5. Riskiest Assumption Test (RAT)

  1. List all assumptions
  2. Rank by: Risk × Importance
  3. Identify #1 riskiest
  4. Design minimal test
  5. Run test, learn
  6. Repeat
  

  6. Evidence Strength

  Evidence	Strength
  Someone paid	Very strong
  Clear commitment	Strong
  Behavior change	Strong
  Stated intent	Medium
  Opinion	Weak
  Assumption	None

• name: Continuous Discovery Habits description: Making discovery a habit, not a phase when_to_use: Establishing ongoing discovery practice implementation: |

  Continuous Discovery

  1. Weekly Cadence

  Minimum Viable Discovery:
  
  Every week:
  - 1 customer touchpoint (interview, observation)
  - Review usage data
  - Update opportunity map
  - Share learnings with team
  
  Time: ~3-4 hours/week
  

  2. Discovery Types by Week

  Week	Focus
  Week 1	Customer interviews (problem)
  Week 2	Prototype testing (solution)
  Week 3	Data analysis (behavior)
  Week 4	Synthesis and planning

  3. Discovery Touchpoints

  Interview Opportunities

  • Recent sign-ups
  • Recent churns
  • Power users
  • New segment entrants
  • Support ticket authors

  Observation Opportunities

  • Session recordings
  • Support calls
  • Sales calls
  • Customer meetings

  4. Discovery Rituals

  Weekly Discovery Share

  • 15-minute team meeting
  • "Here's what we learned"
  • "Here's what we're testing"
  • Anyone can share insights

  Monthly Discovery Synthesis

  • Pattern recognition
  • Opportunity map update
  • Priority recalibration

  5. Discovery-Delivery Balance

  Ratio by team phase:
  
  Early product: 70% discovery / 30% delivery
  Growing product: 40% discovery / 60% delivery
  Mature product: 20% discovery / 80% delivery
  
  Never 0% discovery.
  

  6. Discovery Artifacts

  Artifact	Purpose	Update Frequency
  Opportunity map	Visualize problem space	Weekly
  Interview repository	Store learnings	After each interview
  Assumption tracker	Track risks	Weekly
  Customer personas	Understand segments	Monthly

• name: Prototype Testing description: Testing solutions before building when_to_use: Validating solution ideas implementation: |

  Prototype Testing

  1. Prototype Fidelity Levels

  Level	Fidelity	Time to Build	Best For
  Paper	Very low	Minutes	Concept testing
  Wireframe	Low	Hours	Flow testing
  Interactive	Medium	Days	Usability testing
  Functional	High	Weeks	Feasibility + value

  2. What to Test

  Concept Testing (Paper/Verbal)

  • Does this resonate?
  • Do they understand it?
  • Would they want it?

  Usability Testing (Interactive)

  • Can they use it?
  • Where do they struggle?
  • What confuses them?

  Value Testing (Functional)

  • Does it solve the problem?
  • Is it better than current solution?
  • Would they switch?

  3. Prototype Testing Script

  Setup:
  - This is a prototype, not finished
  - Thinking out loud helps
  - No wrong answers
  
  Task:
  - "Imagine you want to [goal]..."
  - "Show me how you'd do that"
  - Observe silently
  
  Debrief:
  - "What worked well?"
  - "What was confusing?"
  - "How does this compare to your current way?"
  

  4. Sample Size

  Test Type	Users Needed	Rationale
  Usability	5	Find 80% of issues
  Concept	8-12	Pattern recognition
  Preference	20+	Reliable comparison

  5. Quick Prototype Tools

  Need	Tools
  Wireframes	Balsamiq, Whimsical
  Interactive	Figma, Framer
  Functional	Webflow, no-code
  Fake door	Landing page

anti_patterns:

• name: Discovery Theater description: Going through motions without learning why_bad: | Wastes time without reducing risk. Creates false confidence. Check-the-box mentality. what_to_do_instead: | Focus on genuine learning. Ask hard questions. Let evidence change your mind.

• name: Confirmation Bias description: Seeking evidence that confirms beliefs why_bad: | Misses disconfirming evidence. Leads to building wrong thing. Wastes resources. what_to_do_instead: | Seek disconfirming evidence. Ask "what would prove us wrong?" Listen for surprises.

• name: Skipping to Solutions description: Jumping to solutions before understanding problems why_bad: | Solves wrong problems. Misses better solutions. Features without value. what_to_do_instead: | Spend time in problem space. Generate multiple solutions. Test before building.

• name: Discovery Phase Ending description: Treating discovery as a phase that ends why_bad: | Customer understanding decays. Market changes missed. Product-market fit erodes. what_to_do_instead: | Continuous discovery habits. Weekly customer touchpoints. Never stop learning.

• name: Only Asking Power Users description: Research limited to existing heavy users why_bad: | Misses acquisition blockers. Biased toward current patterns. Ignores potential customers. what_to_do_instead: | Interview across user lifecycle. Include non-users and churned. Seek diverse perspectives.

handoffs:

• trigger: "prioritization|roadmap|what to build first" to: feature-prioritization context: "Need to prioritize discovered opportunities"

• trigger: "product-market fit|retention|traction" to: product-market-fit context: "Need to validate PMF"

• trigger: "product strategy|vision" to: product-strategy context: "Need strategic product direction"

• trigger: "user experience|usability" to: ux-design context: "Need UX expertise for testing"