identify-assumptions（确定假设）

Assumption Mapping Expert

Overview

Systematically identify, categorize, and prioritize the assumptions underlying your product decisions. This skill extends Teresa Torres' four risk categories with four additional categories for new products, and uses a devil's advocate approach from PM, Designer, and Engineer perspectives to surface hidden assumptions.

When to Use

• After ideation, before committing to build.
• When a product decision "feels right" but has not been validated.
• When the team disagrees on risk or priority -- assumptions make disagreements explicit.
• Before designing experiments -- test the riskiest assumptions first.

Risk Categories

Core 4 Categories (Existing Products)

These four categories come from Teresa Torres' *Continuous Discovery Habits* and cover the primary risks for features within an established product:

Extended 8 Categories (New Products)

For new products, four additional risk categories become critical:

Methodology

Phase 1: Devil's Advocate Assumption Surfacing

For each product idea or decision, adopt three adversarial perspectives:

PM Devil's Advocate "I challenge whether this is worth building."

• Is there real demand, or are we projecting our own preferences?
• Will this move the metric we care about?
• Can we sustain this economically?
• Does this align with strategy, or is it a distraction?

Designer Devil's Advocate "I challenge whether users will actually use this."

• Will users discover this feature?
• Can they complete the task without help?
• Does this add complexity that hurts the overall experience?
• Are we designing for edge cases and assuming they are common?

Engineer Devil's Advocate "I challenge whether we can build and maintain this."

• Do we have the technical skills and infrastructure?
• What are the hidden dependencies and integration risks?
• Can this scale if it succeeds?
• What is the ongoing maintenance burden?

Phase 2: Categorize Each Assumption

For each assumption surfaced, assign:

Phase 3: Prioritize Using Impact x Risk Matrix

Calculate a risk score for each assumption:

Risk Score = Impact x (1 - Confidence)

Where confidence maps to: High = 0.8, Medium = 0.5, Low = 0.2

Phase 4: Classify into Quadrants

Place each assumption on a 2x2 matrix:

                    HIGH IMPACT
                        |
     Proceed with       |     Test Now
     Confidence         |     (highest priority)
                        |
  ──────────────────────┼──────────────────────
                        |
     Defer              |     Investigate
     (low priority)     |     (may be important)
                        |
                    LOW IMPACT

         LOW RISK ◄─────┼─────► HIGH RISK

Phase 5: Suggest Tests

For each "Test Now" assumption, recommend a validation approach:

Python Tool: assumption_tracker.py

Track and prioritize assumptions using the CLI tool:

# Run with demo data
python3 scripts/assumption_tracker.py --demo

# Run with custom input
python3 scripts/assumption_tracker.py input.json

# Output as JSON
python3 scripts/assumption_tracker.py input.json --format json

Input Format

{
  "assumptions": [
    {
      "description": "Users will prefer AI summaries over manual notes",
      "category": "value",
      "confidence": "low",
      "impact": 9
    }
  ]
}

Output

Sorted by risk priority with quadrant classification and suggested actions.

See scripts/assumption_tracker.py for full documentation.

Output Format

Assumption Registry

Action Plan for "Test Now" Assumptions

For each assumption in the Test Now quadrant, document:

• Assumption description
• Why it is high risk
• Suggested validation method
• Owner and timeline

Use assets/assumption_map_template.md for the full template.

Integration with Other Discovery Skills

• Use brainstorm-ideas/ to generate ideas whose assumptions you will map.
• Feed "Test Now" assumptions into brainstorm-experiments/ for experiment design.
• Run pre-mortem/ to catch risks that assumption mapping might miss (especially elephants).

Troubleshooting

Success Criteria

• All product decisions have at least 8-15 explicit assumptions documented before build commitment
• Every "Test Now" assumption has an assigned owner, validation method, and timeline
• Risk scores are calculated consistently using the Impact x (1 - Confidence) formula
• Assumptions are re-scored after each validation experiment (confidence levels update)
• At least 80% of "Test Now" assumptions are validated or invalidated before major build decisions
• Assumption map is reviewed and updated at every product discovery cycle (weekly or bi-weekly)
• The team can articulate the top 3 riskiest assumptions for any active initiative

Scope & Limitations

In Scope:

• Systematic assumption identification using PM/Designer/Engineer devil's advocate perspectives
• 8-category risk classification (Value, Usability, Viability, Feasibility, Ethics, Go-to-Market, Strategy, Team)
• Quantitative risk scoring using Impact x (1 - Confidence) formula
• Quadrant classification (Test Now, Proceed, Investigate, Defer) with suggested validation methods
• Assumption registry with priority sorting and action plan generation

Out of Scope:

• Running validation experiments (see brainstorm-experiments/ skill)
• Product strategy or roadmap decisions (see execution/outcome-roadmap/)
• Technical feasibility deep-dives or architecture reviews (see engineering/ skills)
• Financial modeling for viability assumptions (see finance/ domain skills)

Important Caveats:

• Confidence levels (high/medium/low) map to fixed numeric values (0.8/0.5/0.2). This is a simplification; real confidence is continuous.
• The quadrant threshold for "high impact" is set at 7/10. Adjust this threshold for your risk tolerance.
• Assumption mapping is most effective when done collaboratively (Product Trio), not by a single person.

Integration Points

Tool Reference

assumption_tracker.py

Tracks, scores, and prioritizes product assumptions using an Impact x Risk matrix with quadrant classification.

Input fields per assumption:

• description (required): Clear statement of what must be true
• category (required): One of value, usability, viability, feasibility, ethics, gtm, strategy, team
• confidence (required): One of high, medium, low
• impact (required): Integer 1-10

References

• Teresa Torres, *Continuous Discovery Habits* (2021)
• David J. Bland & Alexander Osterwalder, *Testing Business Ideas* (2019)
• Ash Maurya, *Running Lean* (2012)
• Marty Cagan, *Inspired* (2018)