Why Is My Pipeline Full but Revenue Flat?

A full pipeline with flat revenue has four causes: stage inflation, a lengthening sales cycle, win-rate decay and deal-size mix shift. Each cuts revenue about 21% while your open opportunity count holds or rises. Only a created-date cohort separates them. Ebsta and Pavilion put the 2025 new-logo win rate at 19%.

• The symptom: pipeline value and coverage ratio flat or up, closed-won revenue flat or down, two quarters running.
• The four candidates: stage inflation · lengthening cycle · win-rate decay · deal-size mix shift.
• The test that separates them: group opportunities by the month they were created, not the month they are forecast to close, and read four numbers per cohort at an equal age.
• The benchmark: 19% new-logo win rate, 68% MQL-to-SQL, 91-day new-business cycle, 8 stakeholders per deal (Ebsta x Pavilion, 655,000 opportunities, $48 billion).
• The fastest partial fix: re-weight open deals by how long each has already slipped. Deals slipped three months close at 8%, against 18% for deals slipped a week.

The four causes — and the arithmetic that makes them indistinguishable

Pipeline velocity is the only formula that connects pipeline to revenue: (open opportunities × win rate × average deal value) ÷ average sales cycle in days. Each of the four causes moves exactly one term in it. That is why they are easy to confuse: on a dashboard reporting open count and pipeline value, all four look the same.

Take a US B2B team with 400 open opportunities, a $28,000 average deal, a 19% win rate and a 91-day cycle. Baseline velocity is $23,385 a day, or $2,128,000 a quarter. Now move one term at a time.

The second row is the perverse one. A sales cycle that stretches from 91 days to 115 days makes your open pipeline 26% bigger and your revenue 21% smaller at the same time — because at a steady creation rate, the number of deals sitting open is the creation rate multiplied by how long each one stays open. That is not a paradox. It is the same event counted twice, once as good news.

One honest limit on that row: the $18,504 holds the same 400 deals and stretches them over 115 days, so it is the transition — the quarters while the cycle is lengthening, which is exactly when you notice. Once creation rate and cycle length settle, the open count sits 26% higher and revenue per day returns to where it was. If revenue is still down after two full cycles, the cause is not the cycle.

Why a close-date pipeline report cannot tell them apart

Almost every CRM pipeline report groups opportunities by expected close date, and that grouping is the problem, because a slipping deal moves between groups. When a rep pushes a close date from March to June, the March column shrinks, the June column grows, and the deal enters the forecast twice without ever being re-qualified. Group by close date and you are measuring rep behavior, not buyer behavior.

Worse, most CRMs store one close date per opportunity and overwrite it on edit, so a deal that has already slipped four times leaves no trace unless field-history tracking is switched on for that field. If it is off, turn it on today: you cannot backfill history you never captured. Our breakdown of the 17 measurable sales pipeline stages and what each leak costs sets out which stages have a countable denominator and which are just labels on a card.

The created-cohort cut: the exact query to run

Group opportunities by the month they were created. A creation cohort is fixed: a deal cannot leave the cohort it was born into, so nothing is double-counted and nothing hides by moving right. Read four numbers per cohort, and only compare cohorts of equal age.

SELECT date_trunc('month', created_date)                         AS cohort,
       count(*)                                                  AS created,
       count(*) filter (where stage = 'Closed Won'
                          and close_date - created_date <= 180)  AS won_by_day_180,
       avg(case when is_closed = false
                  or close_date - created_date > 180
                then 1.0 else 0 end)                             AS still_open_at_day_180,
       percentile_cont(0.5) within group (order by close_date - created_date)
         filter (where stage = 'Closed Won'
                   and close_date - created_date <= 180)         AS median_days_to_win,
       percentile_cont(0.5) within group (order by amount)
         filter (where stage = 'Closed Won'
                   and close_date - created_date <= 180)         AS median_won_amount
FROM   opportunity
WHERE  created_date >= current_date - interval '18 months'
  AND  created_date <= current_date - interval '6 months'
GROUP  BY 1 ORDER BY 1;

Two things in that query are the whole method, and both get skipped. The second line of the WHERE clause drops every cohort younger than six months: with a 91-day median cycle, a cohort needs roughly two cycles — 180 days — before its win rate means anything. And every measure is then censored at day 180, so an 18-month-old cohort is read at exactly the same age as a seven-month-old one. The equal-maturity rule: never compare a cohort to one that has had less time to close, and never read an older cohort at its full age. Drop the censor and still-open share falls and win rate rises across the table with nothing driving it but cohort age — which is precisely the trend you are trying to detect. A win rate that looks like it collapsed in June is usually a June cohort that is 70 days old.

The cohort signature: which cause produces which pattern

Run those four numbers across twelve mature cohorts and each cause leaves a different fingerprint. When we audit a flat quarter at LeadsNow, the created-date cohort table is the first artifact we ask a client for, ahead of the pipeline report — because the pipeline report is the thing that hid the problem.

The first two rows are usually the same underlying failure, which is why they are listed first: both come from qualification. Ebsta and Pavilion found 76% of B-player deals lack a critical event — a dated, buyer-owned reason the deal has to close — and that low performers’ deals are 217% more likely to slip at late stage. Win-rate decay and mix shift are real but rarer, and they are the two that survive a pipeline clean-out.

Re-forecast by slippage before you change anything

The fastest correction costs nothing and takes an afternoon: stop applying one flat probability to every open deal and weight each by how long it has already slipped. Ebsta and Pavilion published win rate against slippage duration, and the curve is steep.

At $28,000 a deal, that is $1,610,000 against the $2,128,000 a flat 19% would have forecast — 24% lower, without touching a single deal. Two honest caveats. Those slippage win rates are Ebsta’s whole sample, not your team, so substitute your own base rate and keep the ratios between the bands. And the report states plainly that all percentage figures are relative — its “win rates down 10%” means 19% against roughly 21% the year before, not 19% against 29%. That distinction is miscited constantly.

Before you blame the funnel, check whether your sector moved

Step zero of any flat-revenue diagnosis is external, and in the US it is free. The Census Bureau’s Quarterly Services Survey publishes seasonally adjusted revenue by NAICS sector. In Q2 2026, US professional, scientific and technical services revenue was $816.4 billion, up 1.4% (±1.3%) on Q1 and up 6.4% (±1.7%) year over year.

Read the margin of error before the number. A quarterly change of 1.4% with a ±1.3% error band is statistically almost indistinguishable from no change at all, which is why the year-over-year figure is the one to compare yourself against. If your sector grew 6% and you grew 0%, the cause is internal. If your sector fell and you held flat, your funnel is working and your market is not.

What running this diagnosis actually costs

Honestly: two to four hours the first time, roughly thirty minutes a month after that, assuming a data warehouse or a CRM report builder that supports percentiles. The hard part is not the SQL. It is the three prerequisites almost nobody has: field-history tracking on close date, a written definition of a critical event that every rep applies the same way, and someone with the authority to close out 80 dead opportunities in a week while finance is watching pipeline coverage.

That last one is where most self-serve attempts stop, because closing out dead deals makes this quarter’s pipeline number worse before it makes revenue better. If the cohort cut says the problem is upstream — not enough genuinely qualified conversations entering — the cheapest volume is usually already in your CRM rather than in new ad spend, which is what our US database reactivation service works. If it says conversion on calls you already get, start with how to increase a sales team’s close rate. To weight open deals by model rather than by hand, our note on AI sales forecasting and pipeline accuracy covers what that needs from your data.

One measurement point, because it applies to your before-and-after as much as ours: a revenue lift claim needs a stated window. Ours is defined on the LeadsNow methodology page as trailing three-month closed-deal revenue at month six against the trailing three months before launch, and that page discloses the median is lower than the average. Define your own the same way and a flat quarter becomes arguable rather than atmospheric.

Frequently asked questions

My pipeline is full but nothing is closing — how long before I should worry?

One quarter is noise. With a 91-day median cycle, a creation cohort is not mature enough to read until it is about 180 days old, so a single bad-looking month is usually just a young cohort. Worry when two consecutive mature cohorts show the same signature in the table above.

What is a healthy pipeline coverage ratio if my revenue is flat?

Coverage ratio is unfalsifiable on its own, because the required coverage is simply 1 divided by your win rate. At the 19% new-logo win rate reported in the Ebsta x Pavilion 2025 GTM Benchmarks report (655,000 opportunities, $48 billion analyzed), break-even coverage is 5.3x, not the 3x that gets repeated. If you are running 3x coverage at a 19% win rate you are already forecasting a miss.

How do I tell stage inflation from a genuinely slow quarter?

Count open opportunities with zero logged buyer-initiated activity in the last 30 days. Stage inflation shows up as a rising share of the cohort still open with no activity; a slow quarter shows up as activity continuing while median days-to-win rises. The same report found 44% of the contacts sellers interact with are never recorded in the CRM at all, and 26% of those missing contacts are decision-makers — so check activity capture before you conclude a deal is dead.

Will adding more leads fix flat revenue?

Not on its own, and the benchmark data is blunt about it: lead volume rose 20% and MQL-to-SQL conversion reached 68%, while the new-logo win rate fell over the same period to 19%. More top-of-funnel volume passing through a pipeline with a slippage problem produces a bigger pipeline and the same revenue. Fix the term in the velocity equation that actually moved first.

Is my flat quarter my fault or the market’s?

Check your sector before your funnel. US professional, scientific and technical services revenue was $816.4 billion in Q2 2026, up 6.4% year over year, per the Census Bureau’s Quarterly Services Survey release of September 9, 2026. If your sector grew and you did not, the cause is inside your pipeline. If your sector contracted, hold the diagnosis and re-read it next quarter.

Why did my pipeline shrink after we tightened qualification?

Because the pipeline was carrying deals that were never going to close. Expect open count to fall and win rate to rise in the cohorts created after the change, with revenue unchanged for roughly one full cycle — 91 days on the Ebsta new-business average — before it moves. Judging the change before that window closes is the most common reason teams reverse a qualification tightening that was working.