Only about 1% of marketing-generated leads ever become closed revenue, according to Forrester's B2B Revenue Waterfall research. That compression across five or six funnel stages isn't a sign of a broken sales funnel — it's the nature of multi-stage B2B buying. The real question isn't whether your funnel compresses; it's whether each individual stage converts at a rate that matches what's achievable. Here's the stage-by-stage benchmark data, and why the widest range you'll find — MQL to SQL — isn't actually one number.
The Funnel, Stage by Stage
Benchmark ranges compiled across published B2B and B2B SaaS datasets converge on a consistent pattern, even though individual reports vary in their exact figures:
| Stage | Weak | Average | Strong |
|---|---|---|---|
| Visitor to Lead | <1% | 1.5–2.5% | 3–5% |
| Lead to MQL | <20% | 37–41% | 45%+ |
| MQL to SQL | <15% | 13% (cross-industry) / 32–42% (SaaS, well-run) | 50%+ |
| SQL to Opportunity | <30% | 40–48% | 50%+ |
| Opportunity to Customer | <20% | 31–39% | 40%+ |
Run those average figures through 10,000 visitors and the compression becomes visible immediately: roughly 200 leads, 78 MQLs, 29 SQLs, 13 opportunities, and 4–5 closed customers — about 0.04–0.05% end to end. Small improvements at a single stage compound fast, though: moving MQL-to-SQL from 37% to 45% alone turns that same traffic into 6–7 customers instead of 4–5.
Why MQL-to-SQL Has the Widest Range of Any Stage
This is the stage where benchmark reports disagree most, and it's worth understanding why rather than picking a number at random. First Page Sage's directly-tracked B2B SaaS funnel data shows the cross-industry average sitting at roughly 13% under a strict MQL definition — but for B2B SaaS specifically, well-executed programs average 18–40% depending on channel, with SEO-sourced MQLs converting to SQL at roughly 51% versus around 26% for PPC-sourced ones. Both the 13% and the 32–42% range are accurate; they're describing different things. The 13% figure averages every MQL definition across every funnel maturity level, including companies whose "MQL" is just "downloaded a whitepaper." The higher range describes teams with a properly calibrated MQL definition — one requiring both behavioral engagement and firmographic fit — running a specific, competent channel.
The practical implication: if your MQL-to-SQL rate is in the single digits, the fix usually isn't a sales training problem. It's almost always a lead-qualification definition problem. If simply downloading a whitepaper makes someone an MQL, sales will reject the majority of them, and the rate will always look broken regardless of how good your reps are.
Channel Changes the Number More Than Industry Does
Within a single company, the channel that produced an MQL predicts its downstream conversion rate better than industry benchmarks do. First Page Sage's channel-level data shows SEO-sourced MQLs converting to SQL at roughly double the rate of PPC-sourced ones, and paid social (Meta) typically converts lower still than search-based channels, since social is an interruption channel with lower declared intent than someone actively searching for a solution. A blended MQL-to-SQL rate that looks mediocre can be hiding a strong organic channel dragging up a weak paid-social channel — which is a very different fix than "the sales team needs better follow-up."
Call Tracking Changes Your Real Numbers, Not Just Your Reporting
In high-call industries, a form-fill-only view of the funnel meaningfully understates actual conversion. Ruler Analytics' 2026 data shows 56% of legal conversions and 53% of professional-services conversions happen by phone rather than form submission — meaning a high-call business tracking only web forms is under-reporting its true visitor-to-lead rate by roughly half. Before concluding a funnel stage is underperforming, confirm the measurement actually captures every channel a prospect uses to convert.
Enterprise vs. Mid-Market: Same Funnel, Different Shape
Deal size changes where the funnel is tight and where it's loose. Enterprise deals show meaningfully lower visitor-to-lead and lead-to-MQL rates than mid-market — a natural consequence of longer evaluation cycles and buying committees that can run to a dozen or more stakeholders. But enterprise deals also show higher late-stage conversion once multiple decision-makers actually align, because by the time a deal reaches SQL or opportunity stage in an enterprise motion, it has already survived a much longer internal vetting process than a mid-market deal at the same stage. A lower top-of-funnel rate paired with a larger average deal size isn't necessarily underperformance — it can be the expected shape of that specific buyer segment.
How to Use These Benchmarks
- Define every stage before comparing to any benchmark. A "13% average" and a "40% average" for the same stage name usually reflect different definitions, not different performance.
- Diagnose by stage, not by overall lead-to-close rate. A 0.04% end-to-end rate tells you almost nothing about which of five stages is actually the problem.
- Segment MQL-to-SQL by channel before concluding sales is underperforming. A blended rate can hide one strong channel and one weak one.
- Confirm phone conversions are captured if you're in a high-call industry, before trusting a form-only conversion number.
- Weight deal size against funnel-stage rates for enterprise motions, where a "weak" top-of-funnel number paired with strong late-stage conversion can still be a healthy funnel.
The 1% end-to-end figure from Forrester is a useful reality check, not a target: it tells you compression is structural, so the actual work is finding which single stage, once fixed, moves the whole funnel — not chasing an unrealistic lift across all five at once. The same stage-by-stage logic applies just as directly to recruiting funnels, where apply-to-hire compression follows the same structural pattern.