Home/ Skills/ funnel-conversion-analysis

general funnel-conversion-analysis

funnel-conversion-analysis

This skill should be used when the user asks to "analyze funnel conversion rates", "audit our sales funnel", "where is the funnel leaking", "diagnose funnel drop-off", "improve funnel conversion", "funnel stage analysis", "conversion funnel audit", "find where leads are dropping off", "optimize the sales funnel", or any variation of analyzing and improving stage-by-stage conversion rates in a B2B SaaS sales funnel.

Download .md

Funnel Conversion Analysis

Funnel conversion analysis measures how prospects move from one stage to the next, identifies where they drop off, and quantifies the revenue impact of each drop. Every funnel has one or two stages where disproportionate value is lost. Finding those stages and fixing them is the highest-leverage growth work.

The principle: don't optimize the whole funnel at once. Find the single stage with the worst conversion rate relative to benchmark, fix it, and move to the next. A 10% improvement at the worst stage produces more revenue than a 2% improvement across all stages.

The Standard B2B SaaS Funnel

Funnel stages

Stage	Definition	What enters	What exits
Visitor	Visits the website	Traffic from all sources	Exits site or converts
Lead	Provides contact info	Form fills, sign-ups, chat	Unqualified or moves to MQL
MQL	Meets marketing qualification criteria	Leads that match ICP + engagement threshold	Disqualified or moves to SQL
SQL	Accepted by sales as qualified	MQLs that pass sales qualification (BANT/MEDDPICC)	Rejected or moves to Opportunity
Opportunity	Active deal in pipeline	SQLs with defined next steps	Lost or moves to Closed Won
Closed Won	Deal signed	Opportunities that close	Revenue

Benchmark conversion rates

Stage transition	Benchmark range	Median	Top quartile
Visitor → Lead	1-3%	2%	3%+
Lead → MQL	15-30%	20%	30%+
MQL → SQL	30-50%	40%	50%+
SQL → Opportunity	50-70%	60%	70%+
Opportunity → Closed Won	15-30%	20%	25%+
Lead → Closed Won (overall)	1-3%	1.5%	2.5%+

Running the Analysis

Step 1: Map your actual funnel

1. List every stage in your CRM/tool
2. Define the entry and exit criteria for each stage
3. Identify any custom stages (PQL, SAL, etc.)
4. Confirm the data source for each stage count
5. Confirm the date field used (created date, stage entry date)

Step 2: Pull the numbers

What to pull	Time period	Source
Count of records at each stage	Last 90 days (minimum)	CRM
Stage-to-stage conversion rates	Same period	CRM or BI tool
Average time in each stage	Same period	CRM stage history
Drop-off count at each stage	Same period	Records that never progressed
Revenue at Closed Won	Same period	CRM

Step 3: Calculate conversion rates

Conversion Rate = Records exiting stage / Records entering stage

Example (last 90 days):
  Visitors: 50,000
  Leads: 1,000 (2.0% visitor-to-lead)
  MQLs: 250 (25.0% lead-to-MQL)
  SQLs: 100 (40.0% MQL-to-SQL)
  Opportunities: 60 (60.0% SQL-to-opp)
  Closed Won: 12 (20.0% opp-to-won)

  Overall Lead-to-Won: 1.2%
  Revenue: 12 × $25K ACV = $300K

Step 4: Find the bottleneck

Compare each stage's conversion rate to the benchmark. The stage with the largest gap is your bottleneck.

  Lead → MQL: 25% (benchmark: 20%) → Above benchmark. OK
  MQL → SQL: 40% (benchmark: 40%) → At benchmark. OK
  SQL → Opp: 60% (benchmark: 60%) → At benchmark. OK
  Opp → Won: 12% (benchmark: 20%) → 40% below benchmark. BOTTLENECK

  Revenue impact of fixing opp-to-won:
    Current: 60 opps × 12% = 7.2 wins
    At benchmark: 60 opps × 20% = 12 wins
    Delta: 4.8 wins × $25K = $120K per quarter

Diagnosing Each Stage

Visitor → Lead (low: < 1.5%)

Possible cause	Evidence	Fix
Traffic is unqualified	High bounce rate, low time on site	Tighten channel targeting. Reduce spend on low-intent channels
CTA is weak or hidden	Low form fill rate on high-traffic pages	A/B test CTAs. Move demo button above fold. Reduce form fields
No compelling offer	Visitors browse but don't convert	Add a lead magnet (template, calculator, guide) or make the demo offer more compelling
Page speed issues	High exit rate, especially mobile	Audit page speed. Target < 2 second load

Lead → MQL (low: < 15%)

Possible cause	Evidence	Fix
MQL criteria too strict	Most leads are disqualified by scoring	Review scoring model. Lower thresholds or add behavioral signals
Lead quality is low	Leads don't match ICP firmographics	Tighten upstream targeting. Add qualifying fields to forms
No nurture for cold leads	Leads that don't score immediately are abandoned	Build nurture sequences. Re-engage leads after 30/60/90 days
Enrichment gaps	Can't score leads because data is missing	Add enrichment on form submit. Fill firmographic gaps automatically

MQL → SQL (low: < 30%)

Possible cause	Evidence	Fix
MQL definition is too loose	Sales rejects 50%+ of MQLs	Tighten MQL criteria. Add mandatory ICP fit check
Slow follow-up	MQLs sit in queue for 24+ hours	Implement speed-to-lead. Target < 5 minute response
Poor discovery	SDRs can't convert MQLs to meetings	Train discovery skills. Improve first-call scripts
Routing issues	MQLs go to the wrong rep or sit unassigned	Audit routing rules. Fix gaps in territory/round-robin logic

SQL → Opportunity (low: < 50%)

Possible cause	Evidence	Fix
SQL definition too loose	SQLs pass to AEs but don't have real budget/authority	Tighten SQL criteria. Require confirmed budget and timeline
AE follow-up is slow	Time from SQL to first AE contact > 48 hours	SLA for AE follow-up. Alert if > 24 hours
Handoff breaks	Context lost between SDR and AE	Standardize handoff notes. SDR provides summary before AE call
Wrong persona	SQLs are the right company but wrong person	Add persona qualification to SQL criteria

Opportunity → Closed Won (low: < 15%)

Possible cause	Evidence	Fix
Pipeline bloat	Too many stale opportunities sitting in early stages	Implement deal hygiene. Auto-close opps with no activity > 30 days
Weak discovery	AE doesn't uncover real pain or compelling event	Improve discovery methodology. Mandate MEDDPICC before stage 3
Pricing objections	Deals die at pricing stage	Review pricing positioning. Test different packaging
No champion	Deals stall because nobody internally advocates	Champion development playbook. Identify and build champion before stage 3
Competitor losses	Win rate against specific competitors is low	Build competitive battlecards. Train on competitive positioning

Segmented Analysis

Always segment the funnel by:

Segment	Why	What it reveals
Source (inbound vs outbound)	Different quality and conversion patterns	Outbound may have lower MQL-SQL but higher SQL-Won
Channel (organic, paid, referral)	Channel quality varies	Referral leads may convert 3x better than paid
Segment (SMB, mid-market, enterprise)	Different sales motions	Enterprise has lower conversion but higher ACV
Persona (IC, Manager, VP)	Different buying behaviors	VP leads convert less often but at higher ACV
Time period (month-over-month)	Trends reveal drift	Declining conversion may indicate market shift

Segmentation rules

Blended funnel metrics hide the real story. A 20% opp-to-won rate may be 30% for inbound and 10% for outbound. The fix for each is completely different. Always segment
Compare like to like. Don't compare this quarter's enterprise funnel to last quarter's blended funnel. Same segment, same time period
Minimum sample size: 30 per stage per segment. Below 30, individual deals swing the rate too much. If a segment has fewer than 30 opps, extend the time period or combine segments

Revenue Impact Modeling

Quantifying the fix

For each bottleneck stage, calculate:

Revenue impact = (target conversion - current conversion)
  × stage volume × downstream conversion × ACV

Example: MQL-to-SQL is 30% (target: 40%)
  MQLs per quarter: 250
  Additional SQLs at target: 250 × (0.40 - 0.30) = 25
  SQL-to-Opp: 60% → 15 additional opps
  Opp-to-Won: 20% → 3 additional wins
  ACV: $25K
  Revenue impact: 3 × $25K = $75K per quarter / $300K per year

Measurement

Metric	Definition	Target	Frequency
Stage-to-stage conversion rates	% moving from each stage to next	At or above benchmark per stage	Monthly
Overall lead-to-won rate	End-to-end conversion	1-3%	Monthly
Average time in stage	Days spent in each stage	Segment-dependent	Monthly
Stage drop-off rate	% that never progress from a given stage	Declining over time	Monthly
Revenue impact per stage	$ impact of improving each stage by 10%	Used for prioritization	Quarterly
Funnel velocity	Revenue / average sales cycle length	Increasing over time	Quarterly

Pre-Analysis Checklist

[ ] All funnel stages defined with clear entry/exit criteria
[ ] Data source confirmed for each stage count
[ ] Time period selected (minimum 90 days for meaningful data)
[ ] Stage-to-stage conversion rates calculated
[ ] Each stage compared to benchmark
[ ] Bottleneck identified (largest gap vs benchmark)
[ ] Revenue impact of fixing the bottleneck calculated
[ ] Funnel segmented by at least source and segment
[ ] Sample sizes validated (30+ per stage per segment)
[ ] Historical trend reviewed (is the bottleneck new or chronic?)
[ ] Root causes hypothesized for the bottleneck stage

Anti-Pattern Check

Optimizing every stage simultaneously. You launch 6 initiatives to improve 6 stages at once. Resources are spread thin. None moves the needle. Focus on the single biggest bottleneck. Fix it. Then move to the next
Using blended conversion rates. Overall opp-to-won is 20%. But inbound is 30% and outbound is 8%. The fix for inbound (nurture more) and outbound (qualify harder) are opposite actions. Blended numbers mislead. Segment always
Not quantifying revenue impact. "MQL-to-SQL is below benchmark" is an observation. "Improving MQL-to-SQL from 30% to 40% adds $300K ARR" is a business case. Always calculate the revenue impact
Measuring conversion over too short a period. Last 30 days: 5 MQLs became SQLs out of 20. That's 25%. But 10 of those MQLs entered the funnel 3 days ago and haven't had time to convert. Use cohort-based analysis with enough time for stage progression
Ignoring time-in-stage. Conversion rate is 40% but average time in stage is 45 days. Deals that convert in 15 days win at 60%. Deals that linger past 30 days win at 10%. Time-in-stage is as important as conversion rate
Funnel stages not defined consistently. One rep marks deals as "Opportunity" after the first call. Another waits until a demo is completed. Inconsistent stage definitions produce meaningless conversion rates. Define entry criteria and enforce them
Only looking at conversion rates, not absolute numbers. MQL-to-SQL improved from 30% to 45%. Great. But MQL volume dropped from 250 to 100. Total SQLs went from 75 to 45. Conversion rate up, pipeline down. Track volume alongside rates

Want agents that use skill files like this?

We customize skill files for your brand voice and methodology, then run content agents against them.

Book a call

# Funnel Conversion Analysis

## The Standard B2B SaaS Funnel

### Funnel stages

| Stage | Definition | What enters | What exits |
|-------|-----------|-------------|-----------|
| Visitor | Visits the website | Traffic from all sources | Exits site or converts |
| Lead | Provides contact info | Form fills, sign-ups, chat | Unqualified or moves to MQL |
| MQL | Meets marketing qualification criteria | Leads that match ICP + engagement threshold | Disqualified or moves to SQL |
| SQL | Accepted by sales as qualified | MQLs that pass sales qualification (BANT/MEDDPICC) | Rejected or moves to Opportunity |
| Opportunity | Active deal in pipeline | SQLs with defined next steps | Lost or moves to Closed Won |
| Closed Won | Deal signed | Opportunities that close | Revenue |

### Benchmark conversion rates

| Stage transition | Benchmark range | Median | Top quartile |
|-----------------|----------------|--------|-------------|
| Visitor → Lead | 1-3% | 2% | 3%+ |
| Lead → MQL | 15-30% | 20% | 30%+ |
| MQL → SQL | 30-50% | 40% | 50%+ |
| SQL → Opportunity | 50-70% | 60% | 70%+ |
| Opportunity → Closed Won | 15-30% | 20% | 25%+ |
| Lead → Closed Won (overall) | 1-3% | 1.5% | 2.5%+ |

---

## Running the Analysis

### Step 1: Map your actual funnel

```
1. List every stage in your CRM/tool
2. Define the entry and exit criteria for each stage
3. Identify any custom stages (PQL, SAL, etc.)
4. Confirm the data source for each stage count
5. Confirm the date field used (created date, stage entry date)
```

### Step 2: Pull the numbers

| What to pull | Time period | Source |
|-------------|-------------|-------|
| Count of records at each stage | Last 90 days (minimum) | CRM |
| Stage-to-stage conversion rates | Same period | CRM or BI tool |
| Average time in each stage | Same period | CRM stage history |
| Drop-off count at each stage | Same period | Records that never progressed |
| Revenue at Closed Won | Same period | CRM |

### Step 3: Calculate conversion rates

```
Conversion Rate = Records exiting stage / Records entering stage

Example (last 90 days):
  Visitors: 50,000
  Leads: 1,000 (2.0% visitor-to-lead)
  MQLs: 250 (25.0% lead-to-MQL)
  SQLs: 100 (40.0% MQL-to-SQL)
  Opportunities: 60 (60.0% SQL-to-opp)
  Closed Won: 12 (20.0% opp-to-won)

Overall Lead-to-Won: 1.2%
  Revenue: 12 × $25K ACV = $300K
```

### Step 4: Find the bottleneck

Compare each stage's conversion rate to the benchmark. The stage with the largest gap is your bottleneck.

```
  Lead → MQL: 25% (benchmark: 20%) → Above benchmark. OK
  MQL → SQL: 40% (benchmark: 40%) → At benchmark. OK
  SQL → Opp: 60% (benchmark: 60%) → At benchmark. OK
  Opp → Won: 12% (benchmark: 20%) → 40% below benchmark. BOTTLENECK

Revenue impact of fixing opp-to-won:
    Current: 60 opps × 12% = 7.2 wins
    At benchmark: 60 opps × 20% = 12 wins
    Delta: 4.8 wins × $25K = $120K per quarter
```

---

## Diagnosing Each Stage

### Visitor → Lead (low: < 1.5%)

| Possible cause | Evidence | Fix |
|---------------|----------|-----|
| Traffic is unqualified | High bounce rate, low time on site | Tighten channel targeting. Reduce spend on low-intent channels |
| CTA is weak or hidden | Low form fill rate on high-traffic pages | A/B test CTAs. Move demo button above fold. Reduce form fields |
| No compelling offer | Visitors browse but don't convert | Add a lead magnet (template, calculator, guide) or make the demo offer more compelling |
| Page speed issues | High exit rate, especially mobile | Audit page speed. Target < 2 second load |

### Lead → MQL (low: < 15%)

| Possible cause | Evidence | Fix |
|---------------|----------|-----|
| MQL criteria too strict | Most leads are disqualified by scoring | Review scoring model. Lower thresholds or add behavioral signals |
| Lead quality is low | Leads don't match ICP firmographics | Tighten upstream targeting. Add qualifying fields to forms |
| No nurture for cold leads | Leads that don't score immediately are abandoned | Build nurture sequences. Re-engage leads after 30/60/90 days |
| Enrichment gaps | Can't score leads because data is missing | Add enrichment on form submit. Fill firmographic gaps automatically |

### MQL → SQL (low: < 30%)

| Possible cause | Evidence | Fix |
|---------------|----------|-----|
| MQL definition is too loose | Sales rejects 50%+ of MQLs | Tighten MQL criteria. Add mandatory ICP fit check |
| Slow follow-up | MQLs sit in queue for 24+ hours | Implement speed-to-lead. Target < 5 minute response |
| Poor discovery | SDRs can't convert MQLs to meetings | Train discovery skills. Improve first-call scripts |
| Routing issues | MQLs go to the wrong rep or sit unassigned | Audit routing rules. Fix gaps in territory/round-robin logic |

### SQL → Opportunity (low: < 50%)

| Possible cause | Evidence | Fix |
|---------------|----------|-----|
| SQL definition too loose | SQLs pass to AEs but don't have real budget/authority | Tighten SQL criteria. Require confirmed budget and timeline |
| AE follow-up is slow | Time from SQL to first AE contact > 48 hours | SLA for AE follow-up. Alert if > 24 hours |
| Handoff breaks | Context lost between SDR and AE | Standardize handoff notes. SDR provides summary before AE call |
| Wrong persona | SQLs are the right company but wrong person | Add persona qualification to SQL criteria |

### Opportunity → Closed Won (low: < 15%)

| Possible cause | Evidence | Fix |
|---------------|----------|-----|
| Pipeline bloat | Too many stale opportunities sitting in early stages | Implement deal hygiene. Auto-close opps with no activity > 30 days |
| Weak discovery | AE doesn't uncover real pain or compelling event | Improve discovery methodology. Mandate MEDDPICC before stage 3 |
| Pricing objections | Deals die at pricing stage | Review pricing positioning. Test different packaging |
| No champion | Deals stall because nobody internally advocates | Champion development playbook. Identify and build champion before stage 3 |
| Competitor losses | Win rate against specific competitors is low | Build competitive battlecards. Train on competitive positioning |

---

## Segmented Analysis

### Always segment the funnel by:

| Segment | Why | What it reveals |
|---------|-----|----------------|
| Source (inbound vs outbound) | Different quality and conversion patterns | Outbound may have lower MQL-SQL but higher SQL-Won |
| Channel (organic, paid, referral) | Channel quality varies | Referral leads may convert 3x better than paid |
| Segment (SMB, mid-market, enterprise) | Different sales motions | Enterprise has lower conversion but higher ACV |
| Persona (IC, Manager, VP) | Different buying behaviors | VP leads convert less often but at higher ACV |
| Time period (month-over-month) | Trends reveal drift | Declining conversion may indicate market shift |

### Segmentation rules

- **Blended funnel metrics hide the real story.** A 20% opp-to-won rate may be 30% for inbound and 10% for outbound. The fix for each is completely different. Always segment
- **Compare like to like.** Don't compare this quarter's enterprise funnel to last quarter's blended funnel. Same segment, same time period
- **Minimum sample size: 30 per stage per segment.** Below 30, individual deals swing the rate too much. If a segment has fewer than 30 opps, extend the time period or combine segments

---

## Revenue Impact Modeling

### Quantifying the fix

```
For each bottleneck stage, calculate:

Revenue impact = (target conversion - current conversion)
  × stage volume × downstream conversion × ACV

Example: MQL-to-SQL is 30% (target: 40%)
  MQLs per quarter: 250
  Additional SQLs at target: 250 × (0.40 - 0.30) = 25
  SQL-to-Opp: 60% → 15 additional opps
  Opp-to-Won: 20% → 3 additional wins
  ACV: $25K
  Revenue impact: 3 × $25K = $75K per quarter / $300K per year
```

---

## Measurement

| Metric | Definition | Target | Frequency |
|--------|-----------|--------|-----------|
| Stage-to-stage conversion rates | % moving from each stage to next | At or above benchmark per stage | Monthly |
| Overall lead-to-won rate | End-to-end conversion | 1-3% | Monthly |
| Average time in stage | Days spent in each stage | Segment-dependent | Monthly |
| Stage drop-off rate | % that never progress from a given stage | Declining over time | Monthly |
| Revenue impact per stage | $ impact of improving each stage by 10% | Used for prioritization | Quarterly |
| Funnel velocity | Revenue / average sales cycle length | Increasing over time | Quarterly |

---

## Pre-Analysis Checklist

- [ ] All funnel stages defined with clear entry/exit criteria
- [ ] Data source confirmed for each stage count
- [ ] Time period selected (minimum 90 days for meaningful data)
- [ ] Stage-to-stage conversion rates calculated
- [ ] Each stage compared to benchmark
- [ ] Bottleneck identified (largest gap vs benchmark)
- [ ] Revenue impact of fixing the bottleneck calculated
- [ ] Funnel segmented by at least source and segment
- [ ] Sample sizes validated (30+ per stage per segment)
- [ ] Historical trend reviewed (is the bottleneck new or chronic?)
- [ ] Root causes hypothesized for the bottleneck stage

---

## Anti-Pattern Check

- Optimizing every stage simultaneously. You launch 6 initiatives to improve 6 stages at once. Resources are spread thin. None moves the needle. Focus on the single biggest bottleneck. Fix it. Then move to the next
- Using blended conversion rates. Overall opp-to-won is 20%. But inbound is 30% and outbound is 8%. The fix for inbound (nurture more) and outbound (qualify harder) are opposite actions. Blended numbers mislead. Segment always
- Not quantifying revenue impact. "MQL-to-SQL is below benchmark" is an observation. "Improving MQL-to-SQL from 30% to 40% adds $300K ARR" is a business case. Always calculate the revenue impact
- Measuring conversion over too short a period. Last 30 days: 5 MQLs became SQLs out of 20. That's 25%. But 10 of those MQLs entered the funnel 3 days ago and haven't had time to convert. Use cohort-based analysis with enough time for stage progression
- Ignoring time-in-stage. Conversion rate is 40% but average time in stage is 45 days. Deals that convert in 15 days win at 60%. Deals that linger past 30 days win at 10%. Time-in-stage is as important as conversion rate
- Funnel stages not defined consistently. One rep marks deals as "Opportunity" after the first call. Another waits until a demo is completed. Inconsistent stage definitions produce meaningless conversion rates. Define entry criteria and enforce them
- Only looking at conversion rates, not absolute numbers. MQL-to-SQL improved from 30% to 45%. Great. But MQL volume dropped from 250 to 100. Total SQLs went from 75 to 45. Conversion rate up, pipeline down. Track volume alongside rates