When to Mark a Deal Closed Lost

When should a deal be marked closed lost?

When it has shown no change in stage, close date, or amount for twelve months, or when the buyer has made a decision that removes you.

Two triggers, one automatic and one evidence-based. The automatic one exists because the evidence-based one never fires often enough on its own.

A twelve-month rule is defensible on the underlying behavior of opportunities. When deals are grouped by pattern and each group gets a predicted time-to-close curve, those curves run from one week to eighty weeks, with most of the closing expectation landing before week twelve. Very few groups carry meaningful expectation past week fifty-two. A deal with a year of no movement sits outside the window where its own cohort resolves.

What test decides it?

Test against meaningful movement, which is a change in stage, close date, or amount.

That distinction is the whole rule. Logged calls and emails are easy to produce without a deal advancing, so any hygiene rule built on activity counts gets satisfied within a week of being announced and stops measuring anything real. Stage, close date, and amount are harder to move without consequence, because moving them changes the forecast.

Apply it in bands rather than as a single cliff.

The evidence trigger runs alongside the clock. A buyer who signed with a competitor, cancelled the project, or lost the budget owner produces a closed-lost record that day regardless of age.

Why do reps keep dead deals open?

Because nothing rewards closing one and several things punish it.

An open deal contributes to every pipeline number a rep appears in. It keeps a manager off a follow-up conversation. It preserves the possibility that months of work produced something. Closing it converts effort into a documented loss.

The system reinforces the behavior. Stage-based reporting makes a stale opportunity look identical to a live one, so nobody scanning a list can tell them apart without running the age test explicitly.

The result is consistent in direction even though the size varies by company. Across ORM customers, 10 percent or more of pipeline has gone untouched for a year, and that value is sitting inside every coverage ratio being reported. Whether the ratio means anything after that is the subject of why the 3x pipeline coverage rule is wrong.

Does closing deals as lost hurt the forecast?

It lowers the pipeline number and raises the quality of everything calculated from it.

Coverage falls. That is the correct outcome and it needs explaining before the pass runs rather than defending afterward. A ratio built partly on deals nobody has touched in a year was never predicting the quarter.

Downstream metrics get better immediately. Win rate becomes meaningful once the denominator stops carrying opportunities that never resolved either way. Average cycle length stops being dragged upward by records that sat for years. Forecast accuracy improves because the model is no longer weighting deals that will never produce an outcome.

One caution. Close deals as lost continuously rather than in a single large pass before a board meeting. A one-time cleanup shows as a cliff in the pipeline chart and invites the wrong question.

Should you delete the record instead?

No. Close it as lost with a reason code and keep it.

A closed-lost record tells you which deal patterns never convert, which sources produce them, and which stages they die in. That is training data for any model that predicts close likelihood, and it is the raw material for win-loss analysis.

A deleted record tells you nothing and cannot be recovered. The pipeline looks the same either way, and only one option leaves you able to explain what happened.

The common objection is that a large volume of closed-lost records makes the team look bad. It makes the team look accurate. Losses that are never recorded do not stop being losses.

What loss reasons do you need?

A short list that separates reasons which expire from reasons which do not. Keep the list under eight options. Long lists get answered with whatever sits at the top of the picklist, and the field stops carrying information.

Make the reason mandatory on close and make one free-text sentence mandatory with it. The sentence is where the useful detail lives.

Who makes the call?

The rule makes the call and the manager documents exceptions.

Individual judgment fails here in a predictable direction. The deals hardest to close as lost are the ones with the most effort invested, which are exactly the ones most likely to be stale.

Set it up so the automation flags and the human overrides. Deals past twelve months of no movement close automatically. Deals in the six to twelve month band close unless a manager writes a specific reason with a dated buyer action attached. The word "still working it" is not a reason.

Run the pass on a schedule rather than on demand. Quarterly full passes with continuous automated flagging in between keeps the volume small enough that no single cleanup distorts the chart. For how this feeds the wider forecasting rhythm, see sales forecasting best practices.