How to tell if your pipeline is realistic

A pipeline is realistic when the total it reports is a number you would bet money on. Most are not, and the gap is rarely one large lie. It accumulates from deals that stopped moving months ago, from stages that record hope rather than evidence, and from close dates nobody has revisited since the deal was created.

Check one: deal age against your real cycle

Find the median time from created to won across your last twenty closed deals. That is your cycle. Now list every open deal older than twice that figure.

You do not have to delete anything today. Flag the list and read it. If you cannot name what happens next on a deal, it is not a live deal, it is a hope with a number attached. Stalled deals have their own repair, and it starts with counting them.

Check two: stage against evidence

For every deal in your late stages, name the thing that put it there. A sent proposal. A named budget. A scheduled decision date. A verbal yes.

If the answer is "the last call went well", the deal is in the wrong stage.

The fix is a written entry condition per stage. What a stage means has to be a fact a second person can verify, and the moment a deal earns the next one has to be the same for everybody on the team.

Check three: close dates that keep moving

Sort your open deals by close date and count how many sit in the past.

Then count how many have been pushed more than twice. A date that has moved three times is not a forecast. It is a placeholder that gets refreshed whenever somebody reviews the board.

A rule that works: a close date may move once without comment. The second move needs a reason written on the record. The third means the deal goes back to whatever stage its actual evidence supports.

Check four: the total against what you actually closed

Take the pipeline total from ninety days ago and compare it with what actually closed in those ninety days. If the pipeline held 200,000 and you closed 20,000, your working conversion is ten percent, whatever your stage percentages claim. Apply that ratio to today's total before you plan against it. That ratio is your win rate, and what counts as a good one depends on where your leads come from.

What to do with the result

Qualification is what stops the problem returning. A deal that was never qualified will fail every one of these checks again a quarter from now.

Frequently asked questions

How do you know if a sales pipeline is realistic?

Compare what the pipeline predicted ninety days ago against what actually closed in that period. The ratio is your working conversion rate, measured on your own deals rather than an industry benchmark. Then check deal age against your median cycle length, and look for close dates that have been pushed more than twice. A pipeline is realistic when those three readings agree with each other instead of contradicting.

What percentage of pipeline should close?

There is no correct percentage, because it depends entirely on how early you qualify and what you count as a deal. The useful figure is your own: divide what you closed in a period by what the pipeline held at the start of it. Measure it the same way each quarter. A team that disqualifies early shows a higher rate than one that keeps everything open, and neither approach is wrong.

How long should a deal stay in the pipeline?

Take the median time from created to won across your last twenty closed deals, and treat twice that figure as the point where a deal needs a decision rather than another follow-up. Past three times the median, most teams find the deal is either already lost or waiting on something nobody has named. The specific number matters less than having one, because without it every deal stays open by default.

Why is my pipeline value always wrong?

Usually because three things accumulate at the same time. Deals that died months ago are still open, stages record optimism instead of evidence, and close dates get pushed rather than reconsidered. Each is small on its own deal, and they compound across the whole board. Running all four checks together removes most of the gap in a single afternoon, without any new tooling.

Should you delete old deals or mark them lost?

Mark them lost, with a reason. Deleting destroys your win rate, your cycle length and every comparison you might want later, because all of those figures are calculated from closed deals. A lost deal with a recorded reason is data you can read next quarter. A deleted one is a hole in your history that makes the next forecast harder to check.

How often should you clean the pipeline?

Monthly is enough for most small teams, and it takes under an hour once each stage has a written entry condition. The cadence matters more than the depth of any single pass. A short monthly review keeps the total close to something you would bet on, while one large annual clean-up means eleven months of planning against a number nobody in the room believes.