Reading time: 14 minutes. By the end you will know exactly where 18 to 22 percent of your ARR is leaking, why your current dashboard cannot show it, and the 30-minute mapping exercise you can run on a Saturday to find it.
The $25M founder who was already at $30M and could not see it
Six weeks ago a SaaS founder sat across from me on a Zoom call. $25.4M ARR. 28 percent net new growth. Series B closed. Board happy. Dashboard green.
I asked him one question.
"How much revenue did you earn last year that never showed up in your bookings number?"
He went quiet for nine seconds. Then he said the thing every founder says at this stage. "I'm not sure what you mean."
Here is what I mean. I have generated over 1 million leads for SaaS and AI companies between $5M and $250M ARR. Our leadership team at Agile Growth Labs has executed over $43B in transactions and 70+ rollups in the last 4 years. The pattern we see inside SaaS businesses at the $10M to $50M band is the same every time. A SaaS company in that range is leaking between 18 and 22 percent of the revenue they already earned. They cannot see it because the leak does not appear on the dashboard their CFO built. It appears in the five places no one is looking.
That same founder, six weeks later, has identified $4.9M of recoverable ARR. Not new pipeline. Recovery from inside the existing motion. The recovery framework is the same one a private equity buyer uses on day one of diligence to figure out what your business is actually worth versus what you think it is worth.
This is that framework, applied in reverse, so you can run it yourself before someone else runs it on you.
Why founders cannot see their own leaks
The reason is structural. Your dashboard was built to answer the question your investors ask: are we growing? It was not built to answer the question a PE buyer asks: how much of the revenue you booked did you actually earn, and how much of the revenue you earned did you fail to book?
Those are different questions. A growth dashboard rolls up ARR, NRR, CAC, and pipeline. A diligence dashboard pulls those apart into 14 sub-metrics that show you where the value is leaking out of the funnel between the moment a buyer decides to spend money with you and the moment that money lands in your bank account.
The leaks live in five categories. Every SaaS business between $10M and $50M has all five. The dollar size shifts. The pattern does not.
Leak 1: Win-Rate Erosion (the "no decision" bucket)
Mechanism. You measure win rate as deals won divided by deals lost. A buyer takes a third path and you do not count it. They go to "no decision." Your CRM either closes the opportunity to a generic "lost" reason or it sits in a stalled stage for nine months until someone deletes it. Industry data puts no-decision losses at roughly 40 to 60 percent of all closed lost deals at the $25M ARR band. Your real win rate is lower than your dashboard says.
Anonymized example. A martech platform we audited last quarter reported a 24 percent win rate on Q3 opportunities. We pulled the raw data. 38 percent of closed lost deals had no documented competitor and a final stage activity of "no response after demo." Reclassified, the real win rate against a decision was 41 percent. The leak: 17 percentage points of opportunities they had already qualified, demoed, and quoted were quietly dying after the demo, with no structured re-engage motion to recover them. Dollar value: $3.2M ARR over the trailing 12 months.
Map it in 30 minutes. Open your CRM. Filter closed lost opportunities from the trailing 12 months. Sort by close stage. Count the ones where the loss reason is blank, "no response," "timing," "budget," or "internal priority." Multiply that count by your average ACV. That dollar number is your no-decision leak. It is recoverable because these buyers already wanted what you sell. They got stuck on the buying side, not the selling side.
Leak 2: Expansion Blindness (your customers buying from your competitor what you already sell)
Mechanism. A buyer signs with you for module A. Eight months later they buy module B from your competitor. Your CSM did not know they were shopping. Your account expansion motion is reactive, triggered by usage signals or QBR conversations, not by intent data on the buyer's actual purchase calendar. Industry expansion benchmark at $25M ARR is 110 to 130 percent NRR. Most companies land in the 95 to 110 range. The delta is the leak.
Anonymized example. A vertical SaaS company in healthcare had 142 enterprise accounts. We mapped 47 of them against G2 buyer intent and LinkedIn job posts. 19 of the 47 had publicly hired for a role that mapped exactly to a product the SaaS company already sold. None of those 19 had received an expansion outreach in the prior 90 days. Reclassified leak: $1.8M ARR sitting in accounts that already trusted them.
Map it in 30 minutes. Take your top 30 accounts by ACV. Pull their LinkedIn company pages and search for hires in the last 90 days that map to a job your product solves. Then look up the same 30 accounts on G2 or your category review site and see which ones have left a review for a competitor product in the last 12 months. Every match is a missed expansion. Multiply by your average expansion ACV.
Leak 3: Follow-Up Decay (the leads who responded once and you never came back)
Mechanism. A lead replies to outreach with "circle back in Q3." Your SDR puts a task in the CRM. Your SDR leaves. The task gets archived during the next territory reshuffle. Q3 arrives. Nobody reaches out. The lead buys a competitor product 11 weeks later. This is the easiest leak to find and the most expensive one to keep. SDR turnover in SaaS averages 14 months. The half-life of a manual follow-up task is shorter than that.
Anonymized example. An AI infrastructure company had 4,200 leads with a "future opportunity" tag in their CRM. We pulled the cohort. 71 percent had been tagged more than 180 days prior. Zero had received a structured re-engage sequence. We built a 5-touch re-engage on the top 600 by ICP score. 41 booked discovery calls. 9 closed in the next 60 days at an average ACV of $84K. Reclassified leak: $756K ARR sitting in tags that nobody owned.
Map it in 30 minutes. In your CRM, build one report: leads tagged "nurture," "future," "not ready," or "circle back" that have not had an outbound activity logged in the last 90 days. The number you get is the size of your dormant pipeline. Multiply the top 20 percent of that list by your blended conversion rate and you have a dollar number for the cohort that is most likely to convert if you simply send them a relevant message this week.
Leak 4: Pre-Churn Invisibility (the 90-day signal nobody read)
Mechanism. A customer churns or downgrades. Your CS team marks it "competitive loss" or "budget." The actual root cause sits 90 days earlier in their usage data, support tickets, or login frequency. By the time anyone notices the signal, the contract is in the renewal window and the buyer has already mentally moved on. Industry data: customers who churn give an average of 3.4 measurable warning signals in the 90 days before they cancel. Most CS teams catch one.
Anonymized example. A B2B SaaS we worked with had 9.2 percent gross churn. We pulled the last 40 churned accounts and overlaid three signals: login frequency drop greater than 40 percent month over month, support ticket sentiment turning negative, and primary user role change on LinkedIn. 33 of the 40 churned accounts had at least two of those three signals firing more than 75 days before the cancel notice. None of those 33 had a save play triggered. Reclassified leak: $2.1M ARR that left because the early warning system was not wired up.
Map it in 30 minutes. Take your last 20 churned accounts. For each one, write down the date they cancelled and the date their primary champion's LinkedIn title changed, the date their login frequency dropped, and the date they opened their last support ticket. Count how many of those 20 had a measurable signal more than 60 days before the cancel date. That percentage times your annual churn dollars equals the saveable portion.
Leak 5: Attribution Misrouting (the deals your CFO is funding the wrong channel for)
Mechanism. Marketing sources a deal. Sales touches it three times. The CRM credits the deal to the last-touch channel, which is "outbound." Marketing budget gets cut in the next quarter because "marketing pipeline is down." Outbound budget gets doubled because "outbound is closing." The channel that actually fed the funnel gets starved. Twelve months later, pipeline collapses and nobody knows why.
Anonymized example. A $32M ARR SaaS company we audited credited 71 percent of new logo bookings to outbound in the trailing 12 months. We pulled the first-touch data from their marketing automation tool and matched it against closed-won records. 44 percent of those "outbound" deals had a marketing-sourced first touch within 180 days of the outbound contact. The deal was already warm. The marketing budget had been cut by $1.4M the prior quarter on the assumption that marketing was not producing pipeline. Reclassified leak: the company was on track to under-invest in the channel actually feeding 44 percent of its closed business, which over 18 months would have compounded into a 30 percent pipeline shortfall.
Map it in 30 minutes. Pull your last 50 closed-won deals. For each one, write down the channel your CRM credits the deal to. Then go into your marketing automation tool and find the first recorded touch for that contact. Note the channel of the first touch. Count how many deals have a different first-touch channel than their CRM-credited channel. That percentage times your next-quarter marketing budget decision is the dollar weight of your attribution misrouting.
The 5-Hour Revenue Leak Audit (the framework)
Five leaks. Each one is a 30-minute mapping exercise. Two and a half hours of work to map. Two and a half hours to validate the dollar numbers with your CRM admin and CFO. Five hours total. The companies that run this audit at $25M ARR find an average of $3.8M to $5.6M of recoverable ARR. The recovery itself takes 90 to 180 days depending on which leaks you prioritize.
Here is the order of operations that gives you the fastest payback:
Week 1. Map all five leaks. Calculate dollar numbers. Rank by dollar size and by speed to recover. Two leaks usually account for 70 percent of the total recoverable number. Start there.
Week 2 to 4. Build the recovery motion for the top two leaks. This is usually a re-engage sequence for Leak 3 plus an expansion play for Leak 2, because those two have the shortest payback. The other three feed the longer cycle.
Week 5 to 12. Wire the early warning systems for Leak 4 and the attribution clean-up for Leak 5. These two prevent the leak from refilling once you drain it.
Week 13 to 24. Sales process work on Leak 1, which is the slowest but compounds the most over a 12-month window because it changes the win rate baseline for every future quarter.
The reason this framework works is the same reason it works inside an M&A diligence. You are not creating new demand. You are finding the demand that already exists inside your business and that your current motion is not capturing. Recovery moves the value forward by 60 to 90 days versus a net-new pipeline play, and the CAC on recovered ARR is roughly one-fifth the CAC on cold pipeline because the buyer has already raised their hand.
At $25M ARR, finding $5M of recoverable ARR is not an aggressive number. It is the median. The aggressive number is finding it before the next board meeting and presenting it as a recovery plan instead of a growth gap.
What to do next
If you want to run this audit on your own business, you can. The mapping exercises above are honest and complete. Five hours, two spreadsheets, and a willingness to look at the parts of your CRM your team has been avoiding.
If you want a second set of eyes on the numbers, that is what we do at Agile Growth Labs. We have run this exact audit across SaaS businesses that have grown into nine-figure exits, and we have run it inside M&A diligence on companies that did not. The framework is the same. The output is a single-page Recovery Map that shows you the five dollar numbers, the prioritization order, and the 90-day plan to recover the highest-value leak.
Book a 30-minute strategy call. We will map two of your five leaks live on the call so you walk away with a number you did not have when you logged on. If the dollar number is below $1M, we will tell you and end the call. If it is above $1M, we will show you what the recovery plan looks like.
Or, if you want the cheaper version first, try the $47 AI Snapshot. It returns your top 3 revenue leaks and the exact fix for each, inside 24 hours.
Frequently Asked Questions
How long does the full Revenue Leak Audit take?
Five hours of mapping plus 90 to 180 days of recovery work. The first dollar of recovered ARR typically lands inside the first 30 days, usually from Leak 3 (the dormant follow-up cohort), because that recovery has the shortest cycle.
What size SaaS business is this framework designed for?
The pattern holds from $5M ARR to roughly $250M ARR. Below $5M the leak categories exist but the dollar sizes are too small to justify the audit time. Above $250M the company typically has a dedicated RevOps function running a version of this internally.
Will this framework work for B2B SaaS only, or also for vertical SaaS and infrastructure?
It works for any SaaS or AI business with a measurable sales cycle and a defined buyer. We have applied it across horizontal SaaS, vertical SaaS, AI infrastructure, and developer tools. The mechanism is the same. The signals you read change slightly per category.
What is the difference between this audit and a standard RevOps assessment?
A standard RevOps assessment looks at process, tooling, and reporting. The Revenue Leak Audit looks at dollar recovery. The output of a standard assessment is a list of improvements. The output of this audit is a number, a prioritization, and a 90-day plan.
How much recoverable ARR do most $25M SaaS companies find?
$3.8M to $5.6M, based on the audits we have run at that ARR band. The distribution is not even across the five leaks. Two of the five usually account for 70 percent of the total. Which two depends on your motion (PLG, sales-led, hybrid) and your category maturity.
Can I run this myself instead of hiring outside help?
Yes. The five mapping exercises in this article are complete and honest. If you have a CRM admin who can pull raw exports and a CFO who can validate dollar numbers, you can run the full audit in one working day. The reason most founders hire outside help is speed and the cost of being wrong, not because the framework is hidden.