When revenue falls short, the usual response is to ask for more leads. Sometimes that is right. Often it is not: the real problem is a stage further down the funnel where good leads are lost, and more leads simply pour more water into a leaking bucket. The way to find out is to size the revenue impact of each stage. This article shows a simple method: map the funnel, compare each stage with a realistic benchmark from your own history, calculate the revenue effect of fixing each stage on its own, and then diagnose the cause at the stage that matters most.
Step 1: map the funnel with real numbers
Take a recent period long enough to be representative, such as the last two quarters, and count records at each stage for a cohort: leads created, those that became MQLs, SQLs, opportunities, proposals and closed won deals. Add the average deal size for won deals. Calculate the conversion rate between each pair of stages. If your CRM data is incomplete for a stage, note it, because the measurement gap may itself be part of the problem. Use separate funnels for inbound and outbound if both are material.
Step 2: set a realistic comparison for each stage
For each conversion, decide what a realistic improved rate would be. The best comparison is your own history: your best recent quarter, your best-performing channel or your best-performing rep or team. External benchmarks can help as a sense check, but they mix companies with very different products, prices and definitions, so treat them with caution. The aim is not a perfect target but a plausible one: what would this stage convert at if it worked as well as it has before, or as well as it does in your best segment?
Step 3: calculate the revenue impact of each stage
For each stage in turn, recalculate the funnel with only that stage at its improved rate and every other stage unchanged. The difference in won revenue is the value of fixing that stage. Because conversions multiply through the funnel, an improvement at any stage raises revenue by the same proportion as the stage's own improvement, so the size of the relative gap matters more than where the stage sits. Then consider the effort and time each fix needs. A smaller revenue gain that can be achieved in weeks may be worth doing before a larger one that needs months.
A worked example
This illustrative B2B funnel shows how the stage with the largest gap is not always the one people assume.
| Stage conversion | Current rate | Realistic rate | Won deals if only this stage improves | Change |
|---|---|---|---|---|
| Lead to MQL | 40% | 45% | 22.5 | +12.5% |
| MQL to SQL | 25% | 40% | 32 | +60% |
| SQL to opportunity | 50% | 55% | 22 | +10% |
| Opportunity to won | 20% | 24% | 24 | +20% |
Starting from 1,000 leads, the current funnel produces 20 won deals (1,000 × 40% × 25% × 50% × 20%). Improving MQL to SQL alone adds the most. Figures are illustrative.
Step 4: diagnose the cause at the leaking stage
Once you know which stage matters most, find out why it underperforms. Break the stage down by source, campaign, segment, region, rep and time to follow up. A low MQL-to-SQL rate might come from one campaign producing poor-fit leads, slow follow-up, an unclear handoff process or a sales team that does not trust marketing leads. A low opportunity-to-won rate might reflect pricing, a competitor, weak proposals or poor qualification earlier on. Listen to a sample of calls, read lost deal notes and ask the people working the stage. The numbers tell you where; conversations usually tell you why.
Common causes by stage
Visitor to lead: unclear offer, weak calls to action, slow or confusing forms, wrong traffic. Lead to MQL: poor-fit audiences, loose targeting, forms that attract the wrong people. MQL to SQL: slow response, unclear handoff rules, poor lead routing, low trust between teams. SQL to opportunity: weak discovery, unclear qualification criteria, wrong person at the prospect. Opportunity to proposal: unclear value, missing stakeholders, no next steps agreed. Proposal to won: pricing, competitive losses, risk concerns, procurement delays, poor follow-up. The sales pipeline leak evaluator on this site helps structure this diagnosis.
Step 5: fix, measure and repeat
Choose one or two fixes at the leaking stage, set a target conversion rate and a date to review, and measure the cohort that went through the stage after the change. Avoid changing several stages at once, because you will not know which change worked. Once the stage improves, re-run the analysis: the biggest leak will often have moved to another stage. Repeating this quarterly keeps effort focused on the change that adds the most revenue.
When more leads really is the answer
Sometimes the analysis shows every stage converting at or near its realistic best. In that case, revenue growth does need more volume at the top of the funnel, or larger deals. Even then, check where new leads would come from and whether they would convert at the same rates: new channels and audiences often convert worse than existing ones. Plan for that in the targets rather than assuming the current rates will hold.
Presenting the analysis to leadership
The result of this analysis is often a change of priorities, so present it clearly. Show the current funnel with volumes and rates, then the table of revenue impact by stage, and state the recommendation in one sentence, for example that improving follow-up between MQL and SQL is worth more than doubling lead volume. Include the evidence for the cause at the leaking stage, such as response times by source or a sample of lost deal reasons, and the specific fixes with owners and a review date. Leaders who have been asking for more leads are more easily persuaded by a revenue figure than by conversion rates, so lead with the revenue impact and keep the rates as supporting detail.
Segmenting before concluding
An average conversion rate can hide very different stories. Before deciding which stage leaks most, split the analysis by the dimensions most likely to differ: lead source, campaign, company size, industry, region, product line and rep. A stage that looks average overall may be excellent for one segment and very poor for another. For example, MQL-to-SQL conversion might be strong for leads from search and weak for leads from a content syndication programme, which points to a source problem rather than a sales follow-up problem. Repeat the revenue impact calculation for the largest segments separately, because the best fix may be to stop feeding a segment into the funnel rather than to change a stage for everyone. Segmenting also guards against fixing a problem that only exists in one part of the business.
FAQ
Map conversion rates for each stage, compare each with a realistic rate from your own history, and calculate the revenue gain from improving each stage on its own.
Use them only as a sense check. Your own best quarter, channel or team is a more reliable comparison because definitions and products match.
It varies. Handoff stages such as MQL to SQL are common culprits because of slow follow-up and unclear ownership, but the analysis should decide, not assumption.
Usually not. Fix one or two at the leaking stage so you can measure which change worked, then repeat the analysis.
When every stage already converts near its realistic best. Even then, expect new channels and audiences to convert differently from existing ones.
Start with the stages you can measure reliably, and treat the gaps as the first problem to fix. Even a rough analysis usually shows which stage deserves attention.
Quarterly works for most B2B companies. The biggest leak often moves once one stage is fixed.
Yes. Replace the B2B stages with sessions, product views, add to cart, checkout started and purchase, and size the revenue impact of each step in the same way.
Choose the one that is faster and cheaper to fix, measure the result, then return to the other.
Marketing, sales and RevOps together. Marketing and sales know the causes at their stages, and RevOps can confirm the data is measured consistently across the funnel.