The leak isn't in the tools. It's in the seams between them.
Every company we audited had bought good software. Salesforce, HubSpot or Marketo, an enrichment vendor, a sequencer, a BI tool, a CS platform. None of them were broken. And yet revenue was leaking — quietly, continuously, and in places no single tool owner was accountable for.
That's the pattern. Leakage lives in the handoffs: lead → router → rep, form → CRM → cadence, signal → CS → save. Each tool does its job and passes the baton. The baton gets dropped in the dark space between them, where no dashboard looks and no one's quota is on the line.
When we instrument those seams and put a dollar figure on them, the number is consistently uncomfortable: 5 to 7 distinct leakage zones, together worth $2M–$15M of ARR per year depending on company size and motion.
140 stacks, one instrumentation pass.
This isn't a survey. For each company we connected to the live systems, reconstructed the end-to-end revenue path, and measured drop-off at every handoff against the conversion the rest of the funnel was already proving was possible. The gap between actual and achievable at each seam — priced at the company's real average deal size and win rate — is the leak.
stacks audited
of companies
stack (median)
per company
Every leak sits between two tools — so it shows up in neither tool's dashboard, and on neither owner's number.
Where it actually goes.
Ranked by how often they appeared and how much they cost. Most companies have five of these live right now; the worst-instrumented have all seven.
Speed-to-lead decay
Inbound → repInbound leads that wait hours — sometimes days — for a first touch. Conversion roughly halves after the first five minutes, yet median first-response in the audits was over an hour. High-intent demo requests rot in a queue while intent cools.
Routing & assignment misfires
Lead → ownerRound-robin sends enterprise leads to SMB reps, territories overlap, and leads land on reps who are on PTO or already over capacity. Misrouted leads convert at a fraction of correctly-routed ones — and the misroute is invisible once the record is "assigned."
Form & funnel friction
Visitor → leadEleven-field forms, broken progressive profiling, qualified anonymous traffic that's never identified, and demo flows that ask for a phone number before they show value. The most expensive abandonment happens one field before the submit button.
The dormant database
MQL → nurtureTens of thousands of past leads that went cold and were never re-engaged, plus a contact database decaying at ~2.5% a month as people change jobs. Deliverability quietly erodes, and re-engageable demand sits inert because no one owns "the list."
Attribution blind spots
Spend → pipelineA dark funnel no model can see means budget gets allocated on the channels that report well, not the ones that convert. Mis-spend compounds quarter over quarter — and the leak here is real dollars poured into underperforming channels.
Data fragmentation & duplicates
Tool ↔ toolDuplicate accounts, sync conflicts between CRM and MAP, and contradictory "sources of truth" that make reps distrust the system and work around it. Every duplicate is a split history, a missed signal, and a deal touched twice or not at all.
Renewal & expansion signal loss
Usage → CSThe single most expensive zone. Churn-risk and expansion signals — usage decay, exec turnover, support-tone shifts, hitting a plan ceiling — exist in the product and never reach CS in time to act. Renewals slip and obvious expansions are left on the table.
Add it up.
No company loses the maximum in all seven zones at once. But the ranges stack — and even a conservative read across a typical mid-market stack lands squarely in eight figures of at-risk ARR per year.
This isn't a software-buying problem. It's a coordination problem — and you've already paid for the data that closes every one of these gaps.
Three structural reasons nobody fixes them.
These zones aren't unknown. Most RevOps leaders can name several off the top of their head. They persist anyway — for three structural reasons that no point tool resolves.
No owner at the seam
Each leak sits between two teams' tools. Marketing owns the form, Sales owns the rep, no one owns the handoff — so it never makes a roadmap.
No dashboard sees it
Tools report on what happens inside them. The drop-off between systems is invisible to every BI view the org already trusts.
Humans can't watch 24/7
Most leaks are timing leaks — a five-minute window, an intent spike, a churn signal. Human-paced workflows miss them by design.
That's why the fix isn't another tool with another dashboard — it's an always-on layer that watches the seams, closes the loop, and learns from every outcome. The leaks are a coordination failure; the answer is autonomous coordination.
Six questions that surface most of the leak.
If you can't answer one of these with a number, that's a zone worth instrumenting. Run them past your RevOps lead this week.
- What is our median first-response time on inbound demo requests?Zone 1 · speed-to-lead
- What share of leads is routed to the wrong segment or an unavailable rep?Zone 2 · routing
- How many qualified anonymous visitors do we never identify?Zone 3 · funnel friction
- How many past MQLs haven't been touched in 90+ days?Zone 4 · dormant database
- What percent of closed-won pipeline is fully attributed to a source?Zone 5 · attribution
- How fast does a churn or expansion signal reach a human who can act?Zone 7 · renewal & expansion
Want the full leakage report — with the benchmarks and the close plan?
The report breaks down all seven zones with the conversion benchmarks behind every number, the instrumentation playbook for each seam, and a 90-day sequence to close the highest-value leaks first.