The short answer: stop using the meeting to find out what the numbers are. A pipeline review that opens with a rep reading their deals aloud has already spent its best twenty minutes on data collection. Get the number, the movement and the reasons on a screen everybody can see before the meeting, and use the room for the only thing a room is good at, which is making a call on the deals where reasonable people disagree.
Why most pipeline reviews produce nothing
Three failures, and they compound.
The data is assembled in the meeting. Somebody exports the CRM on Monday morning, pastes it into a sheet, and by Tuesday afternoon two of the deals have moved and one has gone quiet. The meeting then argues about whether the sheet is right instead of about whether the deal is real.
Every deal gets equal time. A review that walks the list top to bottom gives the same three minutes to a deal that is going to close on its own and a deal that is one unanswered email away from dying. Attention should follow money against risk, not alphabetical order.
Nothing is recorded except a number. The team commits a figure, the quarter ends, and nobody can reconstruct what was believed on the day or why. Without that record, the same optimism is available again next quarter at no cost.
The test. If you cancel next week's review and nothing about the quarter changes, it was a status meeting. That is not an argument for canceling it; it is an argument for changing what happens in it.
The three questions a review has to answer
Everything else is decoration.
- What is the number, and who put their name to it? Not a range, not a feeling, and not an average of the optimists and the pessimists. One figure, with the deals behind it and the person who signed for each.
- What moved since the last one, and why? A total that held still can hide a quarter of a million dollars decaying against nothing strengthening. The delta matters more than the level.
- Which deals are we going to do something about before Friday? Three or four, named, with the action and the owner. A review that ends with fifteen action items ends with none.
What has to be true before the meeting
The preparation is the review. If these four things are true when people arrive, the half hour is worth having.
- Every deal's close date is one somebody would defend today. A date that has already passed is not a forecast, it is a note that nobody has looked. This is the single most common thing wrong with a pipeline, and it is mechanical to find.
- The activity is on the record, not in somebody's head. A deal that looks healthy because the rep remembers a good call is a deal nobody else can judge.
- The movement is computed, not remembered. "It feels slower this month" is not reviewable. "Nine deals worth $1.18 million are decaying and nothing is strengthening" is.
- Somebody has already picked the three deals. Walking in without a shortlist guarantees the list gets walked.
This is the part Empire Command is built to do without anyone preparing it: the pipeline, the forecast and the follow-ups stay current from the conversations the team is already having, and the things that need a decision are ranked by the money attached against the minutes it takes to clear them.
Running the thirty minutes
A shape that works, and why each part is there.
Two minutes on the number. Read the committed figure, who signed it, and the change since last time. Nobody defends anything yet.
Fifteen minutes on three deals. Pick them by exposure, not by stage. For each: what has to be true for this to close on the date it says, what evidence do we have that it is true, and what is the next thing that tests it. If the answer to the third is "follow up", the deal is not being worked, it is being watched.
Five minutes on the ones that went quiet. Silence is the most reliable signal in a pipeline and the easiest to explain away. A deal nobody has spoken to in three weeks is a different conversation from a deal in negotiation.
Five minutes on the calls made. Write down what was decided, by whom, and what would change the answer. This is the part that makes the next review better, and it is the part that always gets cut.
A review that ends with a shorter list than it started with has done its job. One that ends with a longer one has moved work around.
The receipt is what makes the next one better
Forecast accuracy is not a talent. It is a feedback loop, and most teams do not have one, because the thing that was committed on the day is not kept anywhere it can be compared with what actually closed. Keep the submission and the loop closes itself: after two quarters a team can see whose calls hold and whose do not, and the conversation stops being about optimism and starts being about calibration.
We wrote about the mechanics of that in why sales forecasts are wrong by Thursday, and about what to demand from any system that claims to do it in the guide to AI sales forecasting software.
What this does not fix
A better review does not create pipeline, and it does not make a bad quarter good. What it does is stop a bad quarter from being a surprise in week eleven, which is the difference between a problem you can do something about and a problem you can only explain.
Questions people ask about this
How often should a pipeline review happen?
Weekly for the team, and monthly one to one. The weekly one is about movement and the three deals that need a decision; the monthly one is about whether a rep's pipeline is the right shape at all. If the weekly review takes longer than half an hour, the preparation is not being done outside it.
Who should be in the room?
The reps who own the deals being discussed and the person who has to sign for the number. Adding anyone else changes what people say. If a deal needs a specialist, bring them in for that deal rather than parking them through the whole meeting.
What do we do about deals nobody wants to talk about?
Put them on the agenda by exposure rather than by volunteer. The deals nobody raises are usually the ones that have gone quiet, and a deal that has been quiet for more than two weeks is the single clearest signal available that it is in trouble. Naming it is not a punishment; leaving it in the forecast is.
Does a tool fix a bad review?
No. What a tool can do is remove the reasons a review goes wrong before it starts: stale dates, uncounted movement, a pipeline nobody has looked at since Monday, and a number that cannot be traced to the deals behind it. The half hour is still yours to run well or badly.