Most B2B pipeline isn't lost to competitors.It's lost to deals that stall.

Pipeline revenue leakage is revenue lost from deals that stall in the sales process and take much longer to close (or never close at all). This isn't necessarily because a competitor won, but instead is often because your sales motion lost momentum and interest petered out when engagement wasn't correctly targeted and applied.

It happens in the space between marketing and sales that no single team fully owns: in how a lead gets handed over, whether it gets worked, and whether anything notices when a deal stops moving. That makes it an operations problem rather than a sales-effort one, which is why working harder rarely fixes it. For a company selling something considered and complex, it is usually the largest single source of revenue lost.

Between 40 and 60 percent of lost B2B deals end in "no decision" rather than in a loss to a competitor. That range comes from Challenger's JOLT research, built on an analysis of 2.5 million recorded sales conversations. Most deals that don't close were not beaten. They stalled.

The longer a deal slips, the less likely it is to ever close, and the decline is steep. Ebsta and Pavilion, analyzing 655,000 opportunities worth 48 billion dollars in pipeline for the 2025 GTM Benchmarks Report, measured win rates against how long a deal had slipped: 18 percent for deals that slipped by a week, 13 percent at one month, 8 percent at three months, 5 percent at six months, and 3 percent beyond that. A deal that stalls is not just late. It is on a measurable slide.

An earlier edition of the same research, covering 3.2 million opportunities from 364 companies, put a harder edge on it: an opportunity left open longer than twice the length of an average sales cycle had just a 3 percent chance of closing. The same analysis found that opportunities extending beyond their optimum cycle were up to 60 percent less likely to close within an additional month, and up to 90 percent less likely within two.

The pattern shows up everywhere it is measured. Forrester, surveying more than 16,000 business buyers for The State of Business Buying 2024, found that 86 percent of B2B purchases stall at some point in the process. On the buyer side, MarketSource reports that 89 percent of B2B buyers experienced a purchase that stalled in the past year.

The forecast rarely catches any of it, because the dates in the CRM are not measurements. Ebsta found that 89 percent of expected close dates fall on the last day of a calendar month, 68 percent are set earlier than the window in which deals of that size actually close, and 17 percent are changed more than three times by more than a week. A date chosen to fit the quarter is not a prediction.

None of these are edge cases. They are ordinary deals that were in motion and stopped.

Deals rarely stall because of price or because a competitor was better. They stall because momentum breaks down in the space between marketing and sales that no single team fully owns. A lead gets handed over and not worked. A champion goes quiet after an internal review and no one notices for weeks. The buying committee, which Forrester now measures at an average of 13 people, loses alignment, and nobody is watching the signals that would catch it early.

Research supports this. Challenger's JOLT research found moderate to high indecision present in 87 percent of deals, and of the deals lost to no decision, 56 percent were lost to fear of getting it wrong rather than to a preference for the status quo. The customer who cannot decide, not the rival vendor, kills most pipeline.

The reflexive fixes, push harder, follow up more, tighten discovery, treat stalling as an effort problem. It usually isn't. It is a structural problem: the system doesn't recognize when a specific deal is going cold, so no one acts until the deal is already dead.

A stalled deal is not "a deal older than 30 days" or any other fixed threshold, because the window in which deals actually close varies enormously by size. Analyzing 3.2 million opportunities, Ebsta found the period when a deal is most likely to close runs 31 to 60 days for small deals, 61 to 90 days for medium ones, and 150 to 180 days for larger ones. A 60-day-old deal is ancient in one business and perfectly healthy in another.

Stalling is relative to your own cycle. The useful measure is movement, not age: how long a deal has sat without advancing a stage, compared to how long deals like it normally take to move. In the same analysis, opportunities that reached that window had win rates 165 percent higher. The threshold that matters is your own, and it is knowable.

This is also why long-cycle businesses are the most exposed. When deals are supposed to take eight months, a deal sitting silent in month six feels normal, so no one worries, and the deal dies of neglect that looked like patience. The length of the cycle becomes the excuse that hides the stall.

It is tempting to think the cost of a stalled deal is simply the deal you might not win. It is more than that. A stall carries three separate costs, and only the first is obvious.

The first is the lost deal itself. As the win-rate data above shows, a deal that stalls is far less likely to ever close. That is the visible cost.

The second is the work already spent. By the time a deal stalls, your team has usually invested real hours in it: discovery calls, solution design, proposals, pricing, executive sponsor time, and the marketing support behind all of it. When the deal dies quietly, that time is written off, and because it was spread across several people over weeks, no one adds it up. It is real payroll spent on nothing, and it is invisible.

The third is the cost of delay itself, and it applies even to deals that eventually close. A deal that finally lands three months late is not free. The revenue arrives late, so it recognizes late, delivery starts late, services and expansion bill late, and cash flow is worse for the gap. For a business with delivery timelines or services attached, a slow close ties up capacity and pushes revenue into a later period, which is its own quiet loss even when the deal is technically won.

Industry benchmarks can size the first cost. The second and third depend entirely on your own business: your team's time, your delivery economics, your cash position. That is one reason a generic number can only take you so far, and why a real assessment has to look at your actual pipeline.

Because nothing announces it. A stalled deal still sits in the CRM with a dollar amount and a rep assigned. It still appears in the forecast. It looks active right up until the quarter ends and it doesn't close. A system that no one owns end to end does not fail loudly; it leaks quietly, one unworked lead and one silent deal at a time.

That invisibility is why forecast accuracy is such a reliable symptom. SiriusDecisions found that 79 percent of sales organizations miss their forecast by more than 10 percent, and Gartner reports that fewer than half of sales leaders and sellers have high confidence in their forecast at all. The forecast is wrong so often because it is built on pipeline that looks healthier than it is, full of deals that have quietly stopped moving.

You make the invisible visible, then you build the system that catches stalls early enough to act. In practice that means knowing what a normal path looks like for each kind of deal you run, so the system can tell when a specific deal falls off that path, and routing the right response to the right account at the right moment instead of hoping a rep notices.

That is the work Thruvian does: finding the buying patterns already in a company's pipeline and running the engine that keeps deals moving through them, so a stall gets caught in the window where it can still be recovered rather than discovered at quarter close. The generic industry numbers on this page describe the scale of the problem. They cannot tell you where your specific pipeline is leaking. That takes a look at your actual systems.