co-growth-funnel-analysis

name	co-growth-funnel-analysis
description	Funnel analysis skill. AARRR-stage mapping, conversion bottleneck detection, segment-level drop-off analysis, and revenue impact quantification. Use when ANALYZING conversion funnels, diagnosing drop-off points, comparing segment performance, or estimating revenue impact of funnel improvements.

Funnel Analysis

AARRR funnel analysis with bottleneck detection and revenue impact.

Use This Skill When

• Analyzing conversion funnels across AARRR stages.
• Diagnosing where and why users drop off.
• Comparing funnel performance across segments.
• Estimating revenue impact of fixing a specific bottleneck.

Workflow

1. Map user journey to AARRR stages (Acquisition → Referral).
2. Calculate stage-to-stage conversion rates.
3. Identify bottleneck: stage with largest absolute drop-off.
4. Break down drop-off by segment (from user-segmentation output).
5. Quantify revenue impact: (recovered users) × ARPU.
6. Save funnel data, charts, and impact estimates to files.

Deliverables

• report.md: funnel narrative with bottleneck diagnosis.
• results/funnel-metrics.csv: stage, users, conversion rate, drop-off.
• figures/funnel-chart.svg: visual funnel with conversion rates.

Quality Gates

• All 5 AARRR stages are mapped (or gaps explicitly noted).
• Conversion rates are calculated stage-to-stage, not end-to-end only.
• Bottleneck is identified with supporting data.
• Revenue impact estimate includes assumptions.
• Segment-level breakdown is included when segments exist.

If any gate fails: identify the issue, fix, and re-validate.

Gotchas

• End-to-end conversion hides mid-funnel bottlenecks. Always compute stage-to-stage rates.
• Comparing funnels across time periods without controlling for seasonality produces misleading trends.
• "Activation" definition varies by product. Confirm the "aha moment" criteria with the user before analysis.

Validation Loop

1. Generate funnel analysis with metrics
2. Check: all AARRR stages covered, stage-to-stage rates, bottleneck identified
3. If any check fails → revise analysis
4. User approval before prioritizing fixes ⏸️