Pipeline coverage: the honest ratio
Pipeline coverage is the open pipeline you carry, divided by the target for the period. Three to four times is the rule of thumb everyone quotes, and it is exact arithmetic for one shape of company. Yours is probably a different shape, and your own number takes about ten minutes.
By Kshitij Maheshwari, co-founder · Updated August 2026 · 11 min read
Where the number you need actually comes from
Two things set it: the share of your resolved deals you win, and how many times your pipeline turns over inside the period.
| You win | Your cycle, against a quarter | Coverage you need | The shape of company |
|---|---|---|---|
| 4 in 10 | 45 days, half a quarter | 1.25x | Short cycle, one decision-maker |
| 3 in 10 | 45 days, half a quarter | 1.7x | Most seed teams selling to small businesses |
| 1 in 3 | 90 days, the whole quarter | 3x | The one shape the rule of thumb fits |
| 1 in 4 | 90 days, the whole quarter | 4x | Mid-market, a committee of three |
| 3 in 10 | 180 days, two quarters | 6.7x | Enterprise, procurement and a security review |
No published dataset sets a coverage level, so nothing above is a benchmark. Every figure is one division, shown in full further down, and every input is yours: your win rate, your cycle, your period. Find the row nearest yours before you read the rest.
3x is somebody else's win rate, quoted at you
Three times is what winning one deal in three requires. It became a rule when people kept quoting it at companies that win something else.
Pipeline coverage is open pipeline divided by the target for a period. Win rate, here, is wins divided by the deals that actually resolved, won or lost, and not by everything you ever created. Short version in the glossary.
Nobody can point at where three came from. Salesforce's own sales blog ran a piece in 2023 called "Why 3x Pipeline Coverage Is a Terrible Strategy," and the most it will say about the origin is that the saying has been passed down from seller to seller over the years.
Every source that will hand you a coverage benchmark is selling you something that fills a pipeline, and none of them publishes the pipeline it measured.
Three times is exact arithmetic for a company that wins one deal in three. Win four in ten and the same arithmetic says 2.5.
So before you read on, write down what fraction of your resolved deals you actually won. That is the number the rest of this page divides by.
Founders ask us for a coverage target and we do not give one. The number is a fact about your win rate and your cycle, not about outbound. What we can size, once you have those two, is the meeting volume the motion has to produce.
The half of the formula nobody quotes
How many times your pipeline refills inside the period moves the answer as much as your win rate does. Same selling, different number.
Forty-five day cycle, ninety-day quarter
The pipeline turns over twice, so every opportunity gets two chances to land inside the quarter. At a 30% win rate you need 1.7 times the target.
Ninety-day cycle, ninety-day quarter
One turn, one chance per deal. At a 33% win rate you need 3 times, and this single row is the whole of what the famous rule describes.
Six-month cycle, same quarter, same 30%
Half a turn. You need 6.7 times. Nothing about the selling changed, and 3x is now wrong by more than a factor of two, in the expensive direction.
So divide your cycle by your period before you divide by anything else. That one term is the gap between 1.7 and 6.7 at an identical win rate, and it is the term the benchmark tables leave out.
Run it on your last ten closed deals
Four steps on records you already have. You finish holding your own coverage target rather than somebody else's.
Ten minutes, four divisions
4 steps
-
Pull your last ten resolved deals
Won and lost, both. Anything still open is not evidence yet, because it has not told you anything.
-
Divide the wins by those ten
Three won out of ten resolved is 0.3. Dividing by everything you ever created instead is the most common way founders halve their own win rate.
-
Take the median days from first meeting to signature
Across the same ten. Nobody at seed has a sales cycle yet, only a longest deal, and a weak median beats a benchmark because it is about your product.
-
Divide the cycle by the period, then by the win rate
Forty-five days into a ninety-day quarter is 0.5. Divided by 0.3, that is 1.7 times the target. You now have a coverage number, and it is probably not three.
An illustrative walkthrough of the method, not a specific client result. We report real numbers only when they are real.
If your CRM cannot tell you cleanly which deals resolved, fix the counting before the ratio: the six counts and the places each one quietly gets fudged are in our guide to outbound metrics.
One deal moves your answer by half
Ten deals is enough to compute a win rate and not enough to trust one. Both are true, and the fix is not more arithmetic.
Three of ten won needs 3.3 times. Win one more and it is 2.5. Lose one more and it is 5. One deal doubled the pipeline you thought you needed.
That is not a reason to skip the arithmetic. It is a reason to carry the answer as a range, quote it to one decimal at most, and expect it to move. Small numbers mislead in outbound the same way, which is why your A/B test is lying.
Recompute after every fifth resolved deal, not on the first Monday of the month. A review that recalculates the ratio off inputs which have not moved is a meeting about a number that did not move.
Know the meeting number and want the two of us running the motion?
Book a Fit CheckCount deals, not values, while the values are guesses
A seed pipeline's deal values are guesses, and you cannot win 40 percent of a deal. You win it or you lose it.
Weight your guesses and call it a forecast
Acme 60k at 40%, Borel 25k at 60%, Cray 90k at 20%, and three more. Weighted 111k against a 150k target. Coverage 0.7x.
- ✕Most of those values were guessed, not quoted
- ✕The percentages come from stages you named yourself
- ✕Cray alone is 60% of the target and lands whole or not at all
Count the deals, and count the target in deals
Same six opportunities. Three deals needed to hit the quarter. Coverage 2x, in deals.
- ✓You cannot misremember whether a deal exists
- ✓No stage probability to invent
- ✓0.7x says panic, 2x says hold, off one pipeline
Believe the count, because the only thing it asks you to know is which deals exist. Run the ratio in counts this quarter, and switch to value when your last ten deals cluster instead of scattering.
Half your pipeline cannot close in the quarter you are worried about
Coverage counts what is open. It says nothing about whether any of it can finish in time, which is the question you were really asking.
Ebsta and Pavilion, looking at 3.2 million deals across 364 companies in 2022, put it plainly. Deal slippage was running at 37%, and in their words: "The pipeline coverage was there, the opportunities looked healthy, they just couldn't be closed in time." Four tests separate what is open from what can land.
Open longer than twice your average cycle
In that same 2022 data, those deals had a 3% chance of closing. Delete them from the count today, by hand, and do it again next month.
A close date on the last day of a month
89% of close dates in that data sat on the last day of a calendar month. That is a habit, not a forecast. Re-date it off the buyer's own calendar or move it out of the quarter.
Nobody on the buying side has met you in six weeks
No dataset sets this line; it is how we run it. A deal nobody has met about is not closing this quarter. Take it out of the count and back into a sequence.
Steps that physically cannot finish in time
A security review, a procurement queue, a budget that opens in January. Count it against next period's target, where it can actually land, and plan this one without it.
The instinct on a low coverage number is to prospect harder. Deleting is faster, and it moves both sides of the ratio at once: the dead deals were also the ones dragging the win rate down when they finally resolved as losses.
Turning the target into a meeting number
Coverage is where a revenue target becomes an outbound target. Four divisions, four minutes, and almost nobody does them.
-
1
Divide the target by your average deal size
Say it lands on six deals to close this quarter. That is the only step your revenue number appears in, so the rest works in any currency.
-
2
Divide the deals by your win rate
Six deals at three wins in ten means twenty opportunities have to resolve inside the quarter, not merely exist in it.
-
3
Divide the opportunities by your meeting conversion
If two in five held meetings turn into an opportunity, twenty opportunities need fifty held meetings in the quarter.
-
4
Divide by the months in the period
Fifty held meetings across a quarter is about seventeen a month. If you are running five, that gap is the finding, and the arithmetic just made it visible.
An illustrative walkthrough of the method, not a specific client result. We report real numbers only when they are real.
Take that held-meeting number into your sequence plan this week. Everything below a held meeting is already written down: our seed-stage outbound playbook runs held to booked to replies to contacts, and the founder-led sales play is the motion that produces them.
The four ways a healthy coverage number still misses
A coverage number can be arithmetically right and still useless. Each tile below names the thing to go and look at.
Anything open past twice your cycle inflates the ratio and drags down the win rate that sets it. Check: how old is your oldest open deal, today?
Dividing by everything ever created halves the rate and doubles the pipeline you think you need. Check: does your denominator hold only deals that resolved?
Lose it and the quarter is gone whatever the ratio said, because you cannot win part of it. Check: what share of the target is your largest open deal?
The Bridge Group's 2026 account executive research, across 158 B2B companies, put quota attainment at 48%, down from 51% in 2024. More pipeline does not fix a rate. Check: has your win rate moved, or only your pipeline?
A healthy ratio is an expensive kind of comfort
The quarter you miss is usually the quarter the number looked fine in week two. Coverage counts what is open, and what is open includes everything too young, too old or too big to finish in time. The reassurance is why nobody prospected in week three.
What to carry out of this
- 1 Your target is your cycle over your period, divided by your win rate.
- 2 Three times fits one shape: a third won, a cycle as long as the period.
- 3 On ten resolved deals the answer moves by half. Carry it as a range.
- 4 Delete what cannot close in time before you add anything new.
Questions founders ask
What is a good pipeline coverage ratio?
How do I calculate pipeline coverage?
Is the 3x pipeline rule real?
Where did the 3x pipeline rule come from?
What if I have only closed a few deals?
Why do I have 4x coverage and still miss the quarter?
Co-founder of Real Good GTM. He has been the first business hire and Chief of Staff at seed-stage B2B startups, building outbound pipeline before any playbook existed. This post comes from sitting in pipeline reviews where the ratio looked healthy and the quarter was already gone.
Connect on LinkedInTurn the number into a motion
The three pages that pick up where the coverage math stops: the sequence, the counting, and the price of a meeting.
The seed-stage outbound playbook
Held meetings back to booked, replies and contacts, with the stack and the hours it costs.
Read the playbookOutbound metrics that matter
The six counts behind every ratio on this page, and where each one quietly gets fudged.
Read the guideWhat a meeting actually costs
Once you know how many meetings the quarter needs, this is what each one costs to get.
Read the postGot the meeting number? We can run the motion.
Book a fit check. We'll look at the meeting volume your target actually implies, size what signal-based outbound can carry, and tell you straight if outbound is not the right motion for you yet.
Book a Fit CheckNo hard sell. No fake numbers. Real good work speaks for itself.