How do you spot an unreliable pipeline?

You spot an unreliable pipeline by its deals that have been open far longer than your sales cycle, close dates that keep being pushed back or have already passed, deals with no recent activity, and a pipeline that grows while won revenue does not. You check it by counting those signals in your CRM and recalculating the pipeline without the deals that show them. The difference between the figure on the dashboard and what remains is how much air there is in your pipeline.

Why a pipeline looks better than it is

A pipeline grows by itself and only shrinks when someone acts. Creating a new opportunity takes a salesperson a minute and feels like progress. Marking a deal as lost feels like failure and earns nothing. The result is that any pipeline which is not actively maintained will, over time, consist partly of deals that died long ago.

On top of that, the pipeline is often used to judge whether a salesperson is doing enough. A full pipeline then becomes a defence, not a prediction. Why that still does not turn pipeline into revenue is covered in why pipeline is not revenue.

What are the signs of an unreliable pipeline?

Ten signals, and how to check each one.

1. Deals older than twice your sales cycle

Take the median time to close of your won deals over the past year. Count all open deals that have been open for longer than twice that period. Then look at how many such old deals were won in the past. Usually it is a small share. These deals should not count at the normal percentage for their stage.

2. Close dates in the past

An open deal with a close date that has already passed means nobody is looking at it. Count them. A few are normal in the last days of a month. A fifth of your pipeline is a sign that dates carry no meaning.

3. Close dates that keep moving

If your CRM keeps a history of changes, as Salesforce and HubSpot can, count how often the close date of open deals has been pushed back. A deal that has been moved three times rarely closes in the next month it names.

4. The end-of-quarter spike

Plot the close dates of all open deals by week. If there is a large spike in the last week of every quarter or month, the dates were chosen to fall within the period, not because that is when the customer will sign.

5. A pile-up in one stage

A healthy pipeline has a funnel shape: many deals at the front, fewer further on. If a disproportionate share sits in one middle stage, for example "Quote sent", that is often the stage where deals go to die.

6. No activity in the last 30 days

Count the open deals with no logged email, meeting, task or note in the last 30 days. A deal that nothing has happened to for a month is, in most B2B cycles, no longer moving. Adjust the window to your own cycle.

7. Round or identical amounts

Many deals of exactly EUR 10,000 or EUR 25,000 point to placeholders that were never updated. Compare the amounts of won deals with what was invoiced afterwards. If there is a structural difference, the same applies to your open pipeline.

8. Pipeline grows, won revenue does not

Put total pipeline and won revenue per quarter side by side over two years. If the pipeline is clearly growing faster, either your conversion is falling or more dead weight is staying in it. The second is more likely. Speed is a signal here too: how to measure it is explained in pipeline velocity explained.

9. Few lost deals

A win rate of 60 percent sounds good. Then look at how many deals are marked as lost each month. If it is almost none, the win rate is not high: losses are simply not being recorded. You can estimate the real rate by counting old open deals as lost.

10. A handful of deals make up the pipeline

If the three largest deals together are half of your pipeline, your forecast is not a weighted average but a bet on three outcomes. That need not be a bad pipeline, but it is far less certain than the total suggests.

Worked example: how much air is in it

Worked example: suppose your dashboard shows an open pipeline of EUR 3,000,000 for the coming quarter, against a target of EUR 1,000,000. That is 3x coverage, and it feels comfortable.

You apply the filters:

• EUR 900,000 sits in deals that have been open for more than twice your cycle.
• EUR 400,000 sits in deals with a close date in the past and no activity in 30 days.
• EUR 200,000 is duplicate deals.

What remains: EUR 1,500,000. Your real coverage is 1.5x, and if you historically win a third of your good-quality pipeline, you end up at around EUR 500,000. That is half your target, while the dashboard said you had plenty of room. What healthy coverage looks like for your business is covered in pipeline coverage explained.

The amounts are an example. The pattern is what you look for in your own CRM.

What it costs when you miss it

An unreliable pipeline costs money in two ways.

Directly: decisions based on a figure that is too high. You hire people, commit to investments or agree a credit facility on revenue that will not arrive. When the quarter misses, it is too late to correct course.

Indirectly: the deals that are real get less attention. A salesperson with 40 open deals, 25 of them dead, spreads their time across 40. The deals that deserve follow-up do not get it. Stalled deals are also one of the leak patterns described on the use cases page.

A pipeline review that works

A weekly pipeline review that walks through every deal takes a lot of time and yields little. This format puts the time where the difference is.

1. Run the ten checks automatically, for example as a report or list in your CRM, so the review starts with the exceptions and not with the total.
2. Discuss only deals that raise a signal or that sit above an amount threshold.
3. Ask one question per flagged deal: what has the customer done in the last two weeks? Not what the salesperson plans to do, but what the customer did.
4. Decide on the spot: a concrete next step with a date, or lost. No third option.
5. Track how much pipeline disappears each week. If it is structurally zero, deals are not being closed out honestly.

What to do tomorrow

1. Calculate your median sales cycle from last year's won deals.
2. Create a list in your CRM of open deals older than twice that cycle.
3. Create a list of open deals with a close date in the past.
4. Create a list of open deals with no activity in 30 days.
5. Add up the value of the union of those three lists and subtract it from your pipeline.
6. Put the result next to your target. That is the figure you open the next forecast conversation with.

How a reliable pipeline feeds into the revenue forecast is explained in the complete guide to revenue forecasting. If your CRM has bigger gaps than just old deals, forecasting with incomplete CRM data will help.

Frequently asked questions

How often should you clean up the pipeline?

Continuously, in small steps. A weekly list of deals that raise a signal works better than a big clean-up each quarter, because by then that quarter's forecast is already spoilt.

Isn't a large pipeline simply good news?

Only if it is real. A large pipeline full of old deals predicts less revenue than a smaller pipeline of active deals. Size says little without quality.

Should you hold salespeople to account on pipeline quality?

Be careful. If salespeople are measured on pipeline size, the pipeline grows. If they are measured on closing out dead deals, that happens. Reward honesty, not volume.

Which signals matter most?

Age and activity. A deal that has been open much longer than normal and has seen no action for weeks is, in almost every sector, a lost deal that has not yet been called one.