PMF Validation
Three pillars tested with partial correlations, mediation analysis, and open-text evidence.
Scorecard
Priorities
Follow-Through
Energy
Effect Sizes (Forest Plot)
Each dot = mean difference between low (≤2) and high (≥4) groups. Whiskers = 95% bootstrap CI. Green = CI excludes zero.
Market Bridge
Painkiller customers should score worse on pillar indicators if the framework is valid.
1. Priorities
Priority Difficulty × Plan Completion
ρ =Priority Blockers
2. Follow-Through
Overcommit × Plan Completion
ρ =Strategy Gap: Vitamin vs Painkiller
Where healthy people succeed but struggling people fail → product candidatesPlanning Paradox
External Pressure × Overcommit
ρ =
3. Energy
What Drains People
Mean Overwhelm by Drain Type
Higher = more overwhelmed (1–5) · min N ≥ 3Integration Wedge
d = ( )
Mean overwhelm:
Tool count × overwhelm: ρ =
Cross-Pillar Analysis
Are the three pillars independent problems, or do they cluster?
Independence (bootstrap 95% CI)
Unique Predictor Contribution
Partial ρ controls for the other two predictorsPillar Fingerprint by Work Situation
Which roles struggle most with which pillar? Darker = higher score (worse for all except completion where lower = worse). Read columns to see a role's pain profile; read rows to see which roles struggle most with each dimension.
Pillar Fingerprint by Role
Same analysis, but by role (q1_roles multi-select). People appear in every role they hold. Which functional roles struggle most with which pillar?
Who Hurts Most? (by Work Situation)
100% stacked bars sorted by high pain %. Pain score = (5 − productive) + (5 − QoL), range 0–8. High pain ≥ 6.
Who Will Pay
FA + Integration Pain
FA + Failed Strategies
Triple Overlap
Beachhead by Role
Same frustrated-achiever concentration, but by role (q1_roles). People appear in every role they hold. Which functional roles have the most "willing to pay" respondents?
PMF Readiness by Role
Composite score: frustrated achiever % (35%) + tool/system fix demand (25%) + integration pain (25%) + pain intensity (15%). Frustrated achievers = high pain + want an addressable fix. Higher score = more addressable pain for a productivity tool. People with multiple roles appear in each.
What to Build First
Limitations
- Single-item measures — each predictor is one question, not a validated scale. Partial correlations may overfit.
- Self-report — common method variance inflates correlations.
- Correlational — partial correlations and mediation ≠ causation.
- Convenience sample — not generalizable beyond this sample.
- Small N for subgroups — frustrated achiever and strategy gap analyses subdivide the sample.
- LLM extraction — confidence-filtered ≥ 0.5 but not human-verified.
Strengthening: longitudinal data, behavioral tracking, A/B tests, multi-item scales, larger sample.