Running a Quarterly Pipeline Scrub Without Losing Real Deals

Decide in advance what counts as stale: no logged activity in a set number of days, no response across multiple outreach attempts, or a stated timeline that's already passed without an update. Writing these criteria down before you start reviewing individual deals keeps the exercise consistent and defensible, rather than a manager purging deals based on a gut feeling about which ones "seem" dead.

A deal flagged as stale by the criteria isn't automatically dead, it's a candidate for a rep to make one final, real attempt to confirm status before it gets marked closed lost. This protects genuinely slow-but-alive deals from being purged incorrectly, while still forcing an honest confirmation rather than letting a rep leave a truly dead deal open out of optimism or reluctance to update the number.

A scrub cycle runs in this order:

1. Write down the stale criteria before reviewing any deal, such as no logged activity, no response to outreach, or a passed timeline.
2. Flag deals that meet the criteria as candidates for closing, not as dead.
3. Have each rep make one final, real attempt to confirm the deal's status before anything is closed.
4. Close confirmed dead deals with a specific loss reason instead of a generic no-response label.
5. Compare the coverage ratio before and after, and explain the drop in your next pipeline review.

When a stale deal does get closed, require a real, specific loss reason rather than a generic "no response" catch-all. Was it truly no decision, did the buyer choose a competitor you never heard about, or did priorities genuinely shift internally. This data compounds over several scrubs into a much clearer picture of where deals actually die, which is far more useful than a clean-looking pipeline with no explanation behind the numbers.

For example, suppose a scrub flags a batch of stale deals. After rep confirmation, a few turn out to be alive with a new timeline, so they stay open with a dated next step. The rest are closed, and their loss reasons show that most stalled at the same stage, right after the buyer's finance team joined the conversation. That pattern would have stayed invisible if every deal had been closed as no response. Now the cleanup, the coverage number, and the loss data all point to one specific fix, which is confirming budget authority earlier in the process.

Removing genuinely dead deals will shrink your pipeline coverage ratio, and that's the correct outcome, not a problem to hide. A smaller, more honest pipeline coverage number is more useful to leadership than a larger one padded with deals that were never really alive. Present the before-and-after honestly in your next pipeline review rather than letting the drop go unexplained.

A pipeline scrubbed once a year accumulates a lot of dead weight in between, distorting every forecast built in the months before the next cleanup. Quarterly is frequent enough to keep the data reasonably honest without becoming a constant distraction from actually selling. Put it on the calendar as a recurring exercise, ideally timed just before quarterly forecasting so the numbers going into that conversation are already clean.

Running the scrub the day before a board or leadership meeting, specifically to make the pipeline look cleaner for that presentation, defeats the purpose and can look like the numbers are being managed rather than reported honestly. The scrub should happen on a fixed schedule regardless of who's about to see the results, not timed reactively around whichever meeting is coming up next.

If a scrub always seems to happen right before a big presentation and never in the weeks after, that pattern itself is worth noticing, since it suggests the pipeline is being cleaned up for appearances rather than being kept honest as a matter of course. A consistent quarterly cadence, on the calendar regardless of what's being presented when, avoids that appearance entirely.

If a genuinely large cleanup is needed right before a big presentation because the pipeline was neglected for too long, that's worth naming honestly to whoever's about to see the numbers, rather than presenting the post-scrub figures as if they'd been accurate all along. A brief, honest explanation holds up far better over time than numbers that quietly shift with no context given.