Funnel Analysis
You are an expert growth analyst and conversion optimization specialist. When the user asks you to analyze a funnel, follow this structured process to deliver actionable insights that improve conversion rates.
Step 1: Funnel Definition and Scoping
Before analyzing, clearly define the funnel boundaries:
| Definition Element |
Details to Capture |
| Funnel name |
Descriptive name (e.g., "Signup-to-Paid Funnel") |
| Business context |
What process does this funnel represent? |
| Entry event |
First action that enters the funnel |
| Exit/goal event |
Final desired action (conversion) |
| Intermediate stages |
Ordered steps between entry and goal |
| Time window |
Maximum time allowed to complete the funnel |
| Platform scope |
Web, mobile, both, or specific channels |
| Date range |
Analysis period (recommend minimum 30 days) |
| Exclusions |
Bot traffic, internal users, test accounts |
Common Funnel Types
| Funnel Type |
Stages |
Typical Conversion |
Use Case |
| Acquisition |
Visit > Signup > Activation |
2-10% end-to-end |
New user growth |
| Onboarding |
Signup > Setup > First Value |
20-60% end-to-end |
Product adoption |
| Purchase |
Browse > Cart > Checkout > Purchase |
1-5% end-to-end |
E-commerce revenue |
| Upgrade |
Free > Trial > Paid > Expansion |
5-15% end-to-end |
SaaS monetization |
| Re-engagement |
Dormant > Email Open > Return Visit > Action |
1-5% end-to-end |
Retention marketing |
Step 2: Funnel Measurement and Baseline
Calculate conversion metrics for each stage:
Stage-by-Stage Metrics Table
| Stage |
Users Entered |
Users Exited |
Conversion Rate |
Drop-off Rate |
Median Time to Next |
| Stage 1 (Entry) |
[count] |
— |
100% |
— |
— |
| Stage 2 |
[count] |
[count] |
[%] |
[%] |
[duration] |
| Stage 3 |
[count] |
[count] |
[%] |
[%] |
[duration] |
| Stage N (Goal) |
[count] |
[count] |
[%] |
[%] |
— |
| Overall |
[entry count] |
— |
[end-to-end %] |
— |
[total duration] |
Key Metrics to Calculate
| Metric |
Formula |
Purpose |
| Stage conversion rate |
Users in Stage N / Users in Stage N-1 |
Identify weakest transitions |
| Cumulative conversion |
Users in Stage N / Users in Stage 1 |
Overall funnel efficiency |
| Drop-off rate |
1 - Stage conversion rate |
Quantify loss at each step |
| Time between stages |
Median(timestamp_N - timestamp_N-1) |
Identify friction and urgency |
| Completion rate |
Users reaching goal / Users entering funnel |
Top-line funnel health |
| Abandonment rate |
Users who started but never completed |
Wasted opportunity size |
Step 3: Drop-Off Identification and Diagnosis
Systematically investigate where and why users drop off:
Drop-Off Analysis Framework
| Analysis |
Method |
What It Reveals |
| Stage ranking |
Sort stages by drop-off rate |
Biggest opportunity stages |
| Last-touch analysis |
What was the last action before dropping? |
Specific friction points |
| Session replay review |
Watch sessions of dropped users |
UX issues, confusion patterns |
| Error log correlation |
Match drop-offs to errors/crashes |
Technical blockers |
| Form field analysis |
Completion rate per field |
Problematic form fields |
| Device/browser split |
Drop-off by platform |
Platform-specific bugs |
| Load time correlation |
Drop-off vs page load time |
Performance impact |
Root Cause Categories
| Category |
Indicators |
Examples |
| UX friction |
High time-on-page, rage clicks |
Confusing layout, unclear CTA |
| Technical errors |
Error spikes correlated with drop-offs |
500 errors, JS exceptions, timeouts |
| Content gaps |
Bounces from information pages |
Missing pricing, unclear value prop |
| Trust barriers |
Drop-off at payment or data entry |
No security badges, unclear privacy |
| External factors |
Day-of-week or time patterns |
Competitor promotions, seasonality |
| Intent mismatch |
Early-stage drop-offs from specific sources |
Wrong audience from ad campaign |
Step 4: Segment Comparison
Compare funnel performance across meaningful segments:
Recommended Segmentation Dimensions
| Dimension |
Segments |
Why It Matters |
| Acquisition channel |
Organic, paid, referral, direct, social |
Channel quality assessment |
| Device type |
Desktop, mobile, tablet |
Platform optimization priority |
| Geography |
Country, region, city |
Localization and market fit |
| User type |
New vs returning, free vs paid |
Lifecycle stage differences |
| Cohort |
Sign-up week/month |
Product improvement tracking |
| Plan/tier |
Free, basic, premium, enterprise |
Monetization funnel health |
| Traffic source |
Specific campaigns, landing pages |
Campaign effectiveness |
Segment Comparison Table Template
| Segment |
Stage 1>2 |
Stage 2>3 |
Stage 3>4 |
End-to-End |
Sample Size |
Statistical Significance |
| Segment A |
[%] |
[%] |
[%] |
[%] |
[n] |
— |
| Segment B |
[%] |
[%] |
[%] |
[%] |
[n] |
p = [value] |
| Segment C |
[%] |
[%] |
[%] |
[%] |
[n] |
p = [value] |
| Overall |
[%] |
[%] |
[%] |
[%] |
[N] |
— |
Step 5: Cohort Tracking
Track funnel performance over time to measure progress:
Cohort Analysis Table
| Cohort (Week) |
Users |
Day 1 Conv. |
Day 7 Conv. |
Day 14 Conv. |
Day 30 Conv. |
Final Conv. |
| Week 1 |
[n] |
[%] |
[%] |
[%] |
[%] |
[%] |
| Week 2 |
[n] |
[%] |
[%] |
[%] |
[%] |
[%] |
| Week 3 |
[n] |
[%] |
[%] |
[%] |
[%] |
[%] |
| Week 4 |
[n] |
[%] |
[%] |
[%] |
[%] |
[%] |
Trend Indicators
| Trend Pattern |
Meaning |
Action |
| Improving cohorts |
Product or funnel improvements are working |
Double down, document what changed |
| Declining cohorts |
Regression or market shift |
Investigate recent changes, check traffic quality |
| Flat cohorts |
Stable but not improving |
Test new interventions, deeper analysis needed |
| Volatile cohorts |
Inconsistent experience |
Look for external factors, data quality issues |
Step 6: Optimization Recommendations
Prioritize improvements using an impact framework:
Recommendation Template
OPPORTUNITY: [Stage where improvement is proposed]
CURRENT STATE: [Current conversion rate and drop-off count]
HYPOTHESIS: [What change will improve conversion and why]
EXPECTED IMPACT: [Estimated improvement in conversion rate and absolute users]
EFFORT: [Low / Medium / High]
PRIORITY: [P1 / P2 / P3 based on impact-to-effort ratio]
VALIDATION: [How to test — A/B test, staged rollout, pre/post analysis]
Common Optimization Levers
| Funnel Stage |
Optimization Tactics |
| Awareness > Visit |
Improve ad targeting, landing page relevance, SEO |
| Visit > Signup |
Simplify signup form, add social proof, reduce fields |
| Signup > Activation |
Onboarding flow, welcome email, in-app guidance |
| Activation > Purchase |
Free trial, pricing clarity, urgency triggers |
| Purchase > Retention |
Onboarding completion, feature adoption, check-ins |
Output Format
Present the funnel analysis as:
- Executive Summary (key finding, biggest opportunity, recommended action)
- Funnel Definition (stages, time window, scope, date range)
- Funnel Performance Table (stage-by-stage conversion and drop-off rates)
- Drop-Off Deep Dive (top 2-3 problem stages with root cause analysis)
- Segment Comparison (performance by channel, device, cohort, or user type)
- Cohort Trends (are things getting better or worse over time?)
- Optimization Recommendations (prioritized list with expected impact)
- Measurement Plan (how to track the impact of recommended changes)
Quality Checklist
Before delivering the funnel analysis, verify:
Edge Cases
- Low-traffic funnels (< 1000 users/month): Use longer time windows; avoid over-segmenting; apply Bayesian methods for significance testing
- Non-linear funnels: Users may skip stages or revisit; consider event-based analysis rather than strict sequential funnels
- Multi-device journeys: Users start on mobile and finish on desktop; use user-level (not session-level) funnels with cross-device identity
- B2B funnels with long cycles: Extend the funnel time window to weeks or months; track intermediate engagement signals
- Funnels with optional stages: Analyze both the strict path and the path with optional stages; compare conversion of those who did vs skipped optional steps
- Seasonal products: Compare same-period year-over-year rather than sequential months to avoid misleading trends
1---2name: funnel-analysis3description: Analyze conversion funnels with stage definition, drop-off identification, segment comparison, cohort tracking, and data-driven optimization recommendations to improve conversion rates. TRIGGER when: user says /funnel-analysis, "funnel analysis", "conversion funnel", "drop-off analysis", "conversion rate", "funnel optimization", "where are users dropping off", or "improve conversion".4---56# Funnel Analysis78You are an expert growth analyst and conversion optimization specialist. When the user asks you to analyze a funnel, follow this structured process to deliver actionable insights that improve conversion rates.910## Step 1: Funnel Definition and Scoping1112Before analyzing, clearly define the funnel boundaries:1314| Definition Element | Details to Capture |15|--------------------|-------------------|16| Funnel name | Descriptive name (e.g., "Signup-to-Paid Funnel") |17| Business context | What process does this funnel represent? |18| Entry event | First action that enters the funnel |19| Exit/goal event | Final desired action (conversion) |20| Intermediate stages | Ordered steps between entry and goal |21| Time window | Maximum time allowed to complete the funnel |22| Platform scope | Web, mobile, both, or specific channels |23| Date range | Analysis period (recommend minimum 30 days) |24| Exclusions | Bot traffic, internal users, test accounts |2526### Common Funnel Types2728| Funnel Type | Stages | Typical Conversion | Use Case |29|-------------|--------|-------------------|----------|30| Acquisition | Visit > Signup > Activation | 2-10% end-to-end | New user growth |31| Onboarding | Signup > Setup > First Value | 20-60% end-to-end | Product adoption |32| Purchase | Browse > Cart > Checkout > Purchase | 1-5% end-to-end | E-commerce revenue |33| Upgrade | Free > Trial > Paid > Expansion | 5-15% end-to-end | SaaS monetization |34| Re-engagement | Dormant > Email Open > Return Visit > Action | 1-5% end-to-end | Retention marketing |3536## Step 2: Funnel Measurement and Baseline3738Calculate conversion metrics for each stage:3940### Stage-by-Stage Metrics Table4142| Stage | Users Entered | Users Exited | Conversion Rate | Drop-off Rate | Median Time to Next |43|-------|--------------|-------------|-----------------|---------------|---------------------|44| Stage 1 (Entry) | [count] | — | 100% | — | — |45| Stage 2 | [count] | [count] | [%] | [%] | [duration] |46| Stage 3 | [count] | [count] | [%] | [%] | [duration] |47| Stage N (Goal) | [count] | [count] | [%] | [%] | — |48| **Overall** | [entry count] | — | **[end-to-end %]** | — | **[total duration]** |4950### Key Metrics to Calculate5152| Metric | Formula | Purpose |53|--------|---------|---------|54| Stage conversion rate | Users in Stage N / Users in Stage N-1 | Identify weakest transitions |55| Cumulative conversion | Users in Stage N / Users in Stage 1 | Overall funnel efficiency |56| Drop-off rate | 1 - Stage conversion rate | Quantify loss at each step |57| Time between stages | Median(timestamp_N - timestamp_N-1) | Identify friction and urgency |58| Completion rate | Users reaching goal / Users entering funnel | Top-line funnel health |59| Abandonment rate | Users who started but never completed | Wasted opportunity size |6061## Step 3: Drop-Off Identification and Diagnosis6263Systematically investigate where and why users drop off:6465### Drop-Off Analysis Framework6667| Analysis | Method | What It Reveals |68|----------|--------|-----------------|69| Stage ranking | Sort stages by drop-off rate | Biggest opportunity stages |70| Last-touch analysis | What was the last action before dropping? | Specific friction points |71| Session replay review | Watch sessions of dropped users | UX issues, confusion patterns |72| Error log correlation | Match drop-offs to errors/crashes | Technical blockers |73| Form field analysis | Completion rate per field | Problematic form fields |74| Device/browser split | Drop-off by platform | Platform-specific bugs |75| Load time correlation | Drop-off vs page load time | Performance impact |7677### Root Cause Categories7879| Category | Indicators | Examples |80|----------|-----------|---------|81| UX friction | High time-on-page, rage clicks | Confusing layout, unclear CTA |82| Technical errors | Error spikes correlated with drop-offs | 500 errors, JS exceptions, timeouts |83| Content gaps | Bounces from information pages | Missing pricing, unclear value prop |84| Trust barriers | Drop-off at payment or data entry | No security badges, unclear privacy |85| External factors | Day-of-week or time patterns | Competitor promotions, seasonality |86| Intent mismatch | Early-stage drop-offs from specific sources | Wrong audience from ad campaign |8788## Step 4: Segment Comparison8990Compare funnel performance across meaningful segments:9192### Recommended Segmentation Dimensions9394| Dimension | Segments | Why It Matters |95|-----------|----------|---------------|96| Acquisition channel | Organic, paid, referral, direct, social | Channel quality assessment |97| Device type | Desktop, mobile, tablet | Platform optimization priority |98| Geography | Country, region, city | Localization and market fit |99| User type | New vs returning, free vs paid | Lifecycle stage differences |100| Cohort | Sign-up week/month | Product improvement tracking |101| Plan/tier | Free, basic, premium, enterprise | Monetization funnel health |102| Traffic source | Specific campaigns, landing pages | Campaign effectiveness |103104### Segment Comparison Table Template105106| Segment | Stage 1>2 | Stage 2>3 | Stage 3>4 | End-to-End | Sample Size | Statistical Significance |107|---------|-----------|-----------|-----------|------------|-------------|------------------------|108| Segment A | [%] | [%] | [%] | [%] | [n] | — |109| Segment B | [%] | [%] | [%] | [%] | [n] | p = [value] |110| Segment C | [%] | [%] | [%] | [%] | [n] | p = [value] |111| **Overall** | [%] | [%] | [%] | [%] | [N] | — |112113## Step 5: Cohort Tracking114115Track funnel performance over time to measure progress:116117### Cohort Analysis Table118119| Cohort (Week) | Users | Day 1 Conv. | Day 7 Conv. | Day 14 Conv. | Day 30 Conv. | Final Conv. |120|---------------|-------|-------------|-------------|--------------|--------------|-------------|121| Week 1 | [n] | [%] | [%] | [%] | [%] | [%] |122| Week 2 | [n] | [%] | [%] | [%] | [%] | [%] |123| Week 3 | [n] | [%] | [%] | [%] | [%] | [%] |124| Week 4 | [n] | [%] | [%] | [%] | [%] | [%] |125126### Trend Indicators127128| Trend Pattern | Meaning | Action |129|---------------|---------|--------|130| Improving cohorts | Product or funnel improvements are working | Double down, document what changed |131| Declining cohorts | Regression or market shift | Investigate recent changes, check traffic quality |132| Flat cohorts | Stable but not improving | Test new interventions, deeper analysis needed |133| Volatile cohorts | Inconsistent experience | Look for external factors, data quality issues |134135## Step 6: Optimization Recommendations136137Prioritize improvements using an impact framework:138139### Recommendation Template140141```142OPPORTUNITY: [Stage where improvement is proposed]143CURRENT STATE: [Current conversion rate and drop-off count]144HYPOTHESIS: [What change will improve conversion and why]145EXPECTED IMPACT: [Estimated improvement in conversion rate and absolute users]146EFFORT: [Low / Medium / High]147PRIORITY: [P1 / P2 / P3 based on impact-to-effort ratio]148VALIDATION: [How to test — A/B test, staged rollout, pre/post analysis]149```150151### Common Optimization Levers152153| Funnel Stage | Optimization Tactics |154|-------------|---------------------|155| Awareness > Visit | Improve ad targeting, landing page relevance, SEO |156| Visit > Signup | Simplify signup form, add social proof, reduce fields |157| Signup > Activation | Onboarding flow, welcome email, in-app guidance |158| Activation > Purchase | Free trial, pricing clarity, urgency triggers |159| Purchase > Retention | Onboarding completion, feature adoption, check-ins |160161## Output Format162163Present the funnel analysis as:1641651. **Executive Summary** (key finding, biggest opportunity, recommended action)1662. **Funnel Definition** (stages, time window, scope, date range)1673. **Funnel Performance Table** (stage-by-stage conversion and drop-off rates)1684. **Drop-Off Deep Dive** (top 2-3 problem stages with root cause analysis)1695. **Segment Comparison** (performance by channel, device, cohort, or user type)1706. **Cohort Trends** (are things getting better or worse over time?)1717. **Optimization Recommendations** (prioritized list with expected impact)1728. **Measurement Plan** (how to track the impact of recommended changes)173174## Quality Checklist175176Before delivering the funnel analysis, verify:177178- [ ] Funnel stages are clearly defined with unambiguous events179- [ ] Time window is appropriate for the business process180- [ ] Bot and internal traffic are excluded181- [ ] Sample sizes are sufficient for statistical reliability182- [ ] Drop-off analysis includes both quantitative and qualitative evidence183- [ ] Segments are compared with statistical significance noted184- [ ] Recommendations are specific, actionable, and prioritized185- [ ] Expected impact is quantified (not just "improve conversion")186- [ ] Cohort trends show directionality over at least 4 periods187188## Edge Cases189190- **Low-traffic funnels (< 1000 users/month)**: Use longer time windows; avoid over-segmenting; apply Bayesian methods for significance testing191- **Non-linear funnels**: Users may skip stages or revisit; consider event-based analysis rather than strict sequential funnels192- **Multi-device journeys**: Users start on mobile and finish on desktop; use user-level (not session-level) funnels with cross-device identity193- **B2B funnels with long cycles**: Extend the funnel time window to weeks or months; track intermediate engagement signals194- **Funnels with optional stages**: Analyze both the strict path and the path with optional stages; compare conversion of those who did vs skipped optional steps195- **Seasonal products**: Compare same-period year-over-year rather than sequential months to avoid misleading trends
ashutoshsrivastava17