Framework

How to Calculate Your Revenue Leak (EVE Framework)

Tom Regan·Founder & GTM Strategist, Artemis GTM·9 min read·Updated Jul 13, 2026

Quick Answer

The EVE framework prices a revenue leak in three steps: Exposure, how many opportunities pass through the leaky stage; Variance, the gap between your conversion rate and a healthy benchmark; and Economics, the dollar value of each lost conversion. Multiply the three and you get the annual cost of the leak, which is how Artemis ranks what to fix first.

EVE (Economic Value Estimation) quantifies a revenue leak in dollar terms with one formula. The typical B2B SaaS company between $1M and $50M ARR carries a substantial annual leak, directionally around $1.6M, a figure drawn from the engagements we have audited and industry benchmarks, not a controlled study. The framework was developed by Stephan Liozu for B2B pricing strategy and adapted by Artemis GTM for go-to-market diagnostics.

The EVE formula: Exposure, Variance, Economics

Read the formula as three parts that map to the three-step lead above:

Exposure is the volume that flows through the leaky stage, your Lead Volume.
Variance is the gap between a healthy top-quartile benchmark and your current measured rate, or (Benchmark Rate minus Current Rate).
Economics is the dollar value of each lost conversion, your Deal Size.

Multiply the three: (Benchmark Rate minus Current Rate) times Lead Volume times Deal Size equals Annual Revenue at Risk. Use top-quartile benchmarks for your stage and motion, not the median, so you size the gap between current and best-in-class. The output is a defensible dollar figure for one specific leak. Sum across leaks for total annual revenue at risk.

Worked examples across the leak categories

Same fictional company throughout: B2B SaaS at $12M ARR, 200 monthly inbound leads, $30K average deal size, 20% close rate. Every figure below is an illustrative worked example, not a guarantee. Each maps onto a leak in the 7 revenue leaks.

1. Anonymous traffic (leak: anonymous website traffic)

Inputs: 10K monthly visitors, 2% form-fill rate, up to 65% company-level visitor identification (vendor-reported), 25% high-intent rate, 5% conversion to meeting. Modeled: (10K times 65% times 25% times 5% times $30K times 20%) minus (10K times 2% times 5% times $30K times 20%) is roughly $420K per year. Recovers when visitor identification and speed-to-lead automation are wired together.

2. Slow lead response (leak: slow lead response)

Inputs: 200 monthly leads, current sub-5-minute rate 30% versus a 90% benchmark, 21% conversion uplift coefficient. Modeled: 200 times (90% minus 30%) times 21% times 20% times $30K times 12 is roughly $1.81M per year. Often the largest single leak for $5M to $25M ARR companies, and it recovers in 2 to 4 weeks of automation work.

3. MQL-to-SQL handoff (leak: leaky MQL-to-SQL handoff)

Inputs: 200 MQLs per month, 23% benchmark MQL-to-SQL conversion, current 12%, 30% post-SQL close rate, $30K deal size. Modeled: 200 times (23% minus 12%) times 30% times $30K times 12 is roughly $2.38M per year. Usually a process problem (no shared MQL-to-SQL definition, no SLA on first sales touch), not a tooling problem.

4. Low outbound reply rates (leak: low outbound reply rates)

Inputs: 5K outbound touches per month, current reply rate 1.5% versus a 5% signal-based benchmark, 30% reply-to-meeting, 20% close. Modeled: 5K times (5% minus 1.5%) times 30% times 20% times $30K times 12 is roughly $378K per year. Recovers by switching outbound from firmographic-only to signal-based.

5. Win-rate drag (recovers via qualification)

Inputs: 50 late-stage opportunities per month, 28% top-quartile close versus current 18%, $30K deal size. Modeled: 50 times (28% minus 18%) times $30K times 12 is roughly $1.80M per year. These are the deals that should have been disqualified earlier; they recover via better discovery and qualification, which is where the leaky handoff and qualification automation come in, not via more deals.

Prioritize by ROI, not raw dollar impact

The biggest dollar leak is rarely the highest-priority fix. Rank each leak by Priority equals Annual Revenue at Risk divided by (Weeks to Fix times Implementation Cost). Speed-to-lead automation might recover a large leak in three weeks at low cost, scoring far higher than a win-rate fix that recovers a similar amount but takes months of coaching. Sort by ROI per fix-week.

Common mistakes

Using median benchmarks instead of top quartile. EVE measures the gap between current and best-in-class, not current and average.
Mixing direct loss with opportunity cost. EVE is about money that should have closed and did not.
Counting non-ICP volume. If 40% of inbound is not ICP-fit, apply EVE only to the 60% that is, or you claim a leak on leads that would not have closed at any rate.

Frequently asked questions

What is the EVE framework for calculating revenue leak?

EVE (Economic Value Estimation) quantifies a revenue leak by comparing your current performance metric against the benchmark for your stage and motion, multiplied by the volume passing through that stage and the average deal size. Formula: (Benchmark Rate minus Current Rate) times Lead Volume times Deal Size equals Annual Revenue at Risk. The framework was originally developed by Stephan Liozu for B2B pricing strategy and adapted by Artemis GTM for go-to-market diagnostics.

How do I calculate the dollar impact of slow lead response?

Use this formula: Lost MQLs per month equals Monthly leads times (1 minus sub-5-minute response rate) times a conversion uplift coefficient (around 0.21 for B2B SaaS). Multiply by close rate, deal size, and 12 months for annual impact. For 200 monthly leads at a 30% sub-5-minute rate, $30K deal size, and 20% close rate, the modeled figure is roughly $2.1M in annual revenue at risk. This is an illustrative worked example, not a guarantee.

What's the median revenue leak in B2B SaaS?

Roughly $1.6M annually for a B2B SaaS company between $1M and $50M ARR, a directional figure drawn from the engagements we have audited and industry benchmarks, not a controlled study. The largest single leaks in the median company tend to be anonymous website traffic, slow lead response, and a leaky MQL-to-SQL handoff.

How accurate is the EVE methodology?

Direction is reliable; magnitude is estimated within roughly 20 to 30% for any single leak. EVE is a quantification framework, not a prediction model. It gives you a defensible dollar figure to prioritize fixes against, not a forecast. Accuracy depends on trustworthy benchmark data for your stage, clean current-state metrics from your CRM, and honest segmentation of which leads actually belong to your ICP.

Should I include opportunity cost or just direct loss?

Direct loss only. EVE quantifies revenue not captured given current pipeline volume, money that should have closed and did not. Opportunity cost (revenue you might have generated with a different strategy) is a forecasting question, not a leak calculation. Mixing them produces inflated numbers that get dismissed by CFOs.

How do I prioritize which leak to fix first?

Rank by Impact divided by (Time-to-Fix times Implementation Cost). The biggest dollar leak is rarely the highest-priority fix. Speed-to-lead automation might recover a large leak in a few weeks at low cost (high ROI), while win-rate drag might recover a similar amount but require months of coaching (low ROI). Sort by ROI per fix-week, not by raw dollar impact.

Run this play in your own stack

Read the guide, then install the engine.

The Artemis AI GTM Engineer runs a free audit inside your first session, prices each leak in dollars, and builds the fix with you inside your own Claude. See how an agent installs and buys, or start with the free audit that prices all seven leaks.