Blog7 min

How to clean up a messy sales pipeline

Clean a pipeline in one pass rather than continuously: sort by last meaningful change, kill anything that has not moved in twice your sales cycle, force a decision on every deal still open in a closed month, and only then fix fields. The reason it rots is structural — marking a deal lost is an admission and leaving it open costs nothing — so the cleanup only holds if you change that incentive.

Why pipelines rot in the same way every time

Nobody decides to let a pipeline go stale. It happens because of an asymmetry: closing a deal as lost is a visible negative that shows up in a rep's numbers, while leaving it open is invisible and preserves the possibility it comes back. Multiply that across a team and a year, and the open column fills with deals that everyone privately knows are dead.

The second cause is close dates. When a deal slips, the path of least resistance is pushing the date to the end of the next quarter. Do that twice and you have a deal that has been "closing next quarter" for nine months, inflating coverage in every period it touches.

A messy pipeline is not a discipline problem, it is an incentive problem. If the only cost of leaving a dead deal open is tidiness, it will stay open.

The one-pass cleanup

1. Sort by last meaningful change, not by last modified. System timestamps bump on workflow enrolments and integration syncs, so they flag deals as alive that nothing human has touched. Use the date of the last logged call, meeting or note instead.

2. Close anything older than two sales cycles. If your average cycle is 60 days, anything untouched for 120 goes to lost. Not "nurture", not a reduced probability. Lost. It can be reopened if it comes back, and reopening is cheap; carrying it is not.

3. Force a decision on overdue deals. Every deal still open in a month that has already ended gets one of three outcomes: it closed, it was lost, or it moved to a specific new month with a reason. There is no fourth option and no leaving it.

4. Only now fix fields. Values, owners, odds. Doing this first is the classic mistake — you spend an afternoon correcting data on deals you are about to delete.

Four rules that keep it clean

  • A close date can only move once. The second slip closes the deal as lost. This single rule removes most of the rot, because it makes the free option expensive.
  • Lost is not a judgement. Say so out loud and repeatedly. Teams that treat lost deals as failures get pipelines full of zombies; teams that treat them as information get accurate ones.
  • Review the tail monthly, not weekly. Weekly attention on forty deals teaches people to invent movement. Monthly is enough for the long tail and leaves the weekly meeting for the few deals that decide the quarter.
  • Require five fields, not twenty. Every mandatory field is a tax paid by someone mid-sale, and the result is not blanks — it is plausible defaults. A pipeline where every deal sits at 50% looks clean and contains nothing.

Do not clean up in order to forecast

The common trap is treating cleanup as a prerequisite for the forecast, and spending half a day on it before every review. It is not. A forecast needs the handful of deals that move the number, and you can get those from the reps in the room in fifteen minutes.

Hygiene matters for coverage reporting, for pipeline generation targets, and for knowing what your team is actually working on. It does not need to be perfect for a forecast, and waiting for perfect is how reviews end up ninety minutes long and held on stale data anyway.

All articles