The last three weeks are not for closing.

Here is the answer, and nobody likes it. In the last three weeks of a quarter, a good sales team should be doing almost exactly what it was doing in week two.

That is the whole discipline. Most of what else happens in those three weeks is theater with a scoreboard behind it.

What should a sales team actually do in the last three weeks of a quarter?

I have run the week-eleven forecast call more times than I can count, across a national team, and it has a shape you could set your watch by.

The same three deals get recommitted. Somebody has a verbal. Somebody is waiting on legal. A deal that was clearly next quarter back in July is suddenly closing Thursday. And prospecting stops, quietly, because prospecting does not help this quarter and everybody in the room knows it and nobody says it out loud.

Then the quarter lands. And it lands almost exactly where it was always going to land, because a deal that needed six weeks of work in week eleven still needs six weeks of work in week thirteen. You cannot compress a buyer’s calendar by wanting to.

Here is the part that costs real money, and it is not the deal you missed.

The three weeks you spent saving this quarter were the three weeks that were supposed to build the next one. So you close Q3 at whatever number Q3 was always going to be, and then you open Q4 three weeks behind. Do that four times a year and you are running a business that is never once at full speed.

A quarter you have to save was decided in week four.

Andrew S. Grove, High Output Management, 1983

Grove wrote that about building semiconductors, and it is still the cleanest description of the quarter-close problem I have found anywhere in sales.

When a leader spends the last three weeks of a quarter personally working deals, the organization’s output for those three weeks is one person’s output. That is not leverage. That is a very expensive individual contributor with a title.

1. Freeze the commit list in week ten.

Nothing new enters commit after week ten. Not a verbal. Not “he told me it is done.” If it was not committed in week ten, it is next quarter’s deal and you plan like it is.

This sounds like it costs you deals. It does not. It costs you the illusion of deals, which is a different thing and a much cheaper one. The deals that were genuinely going to close in weeks eleven through thirteen were already sitting in commit in week ten. What the freeze kills is the late entry, the one that gets committed under pressure, pulls four people onto it, and slips anyway.

The freeze also does something quieter and better. It makes week ten the real deadline, which moves the whole quarter’s work up three weeks. Reps stop treating week thirteen as the finish line and start treating it as the paperwork.

2. Give each of your three meetings one job, and stop letting them all ask the same question.

Most teams run a pipeline review, a forecast call, and a one-on-one. And all three quietly collapse into the identical question: is this deal going to close. Three meetings, one question, hours a month, and no new information after the first hour.

They are supposed to inspect three different things.

The pipeline review inspects next quarter. Not this one. Coverage, entry points, what is being built. If your pipeline review in week eleven is about week thirteen, you do not have a pipeline review. You have a second forecast call wearing a different name.

The forecast call inspects change only. What moved since last week, and why. If nothing moved, that is the answer and the deal takes ninety seconds. A forecast call is not the place to re-argue a deal’s merits. It is the place to find out what is different.

The one-on-one inspects behavior. Not the number. What the rep did, what they would do again, what they are working on installing. The number is an outcome, and at that altitude outcomes are noise.

Split those three cleanly and you get most of your last three weeks back without adding one hour to anybody’s calendar.

3. Third recommit, and the deal moves out.

Give every committed deal a strike count. A deal that has been committed three weeks running with no date change on the customer’s side is not a commit. It is a hope with a history.

Move it. Out loud, in the meeting, without a fight, and using the same words every single time: this is the third week, it goes to next quarter, and if it closes early we will all be delighted.

This is exactly where single-deal exposure gets dangerous. When one deal carries more than 40 percent of a quarter, it fails to close about four times out of five. And the deal that gets recommitted three times is almost always the big one, because nobody wants to be the person who moved the deal the whole number is standing on. So the quarter’s largest exposure ends up protected by the precise social pressure that should be inspecting it hardest.

Moving it is not pessimism. It is the only way to find out what your number actually is while there is still time to do something about it. That is forecast survivability, not forecast accuracy: not did you call the number, but does the number hold when one thing goes wrong.

4. Do not let the close weeks eat the prospecting block.

This is the one that compounds, which is exactly why it is the first thing cut.

Count where your deals enter. Expansion or add-on business inside an account you already serve closes about 3 in 4. A referral closes about 1 in 5. Cold outbound is closer to 1 in 100. Those are not small differences, and the entry point gets decided weeks before anybody touches a forecast. That is pipeline coverage quality, and it is a completely different question from pipeline coverage quantity.

Which means the three highest-value hours in week twelve are not on a closing call. They are the expansion conversation inside an account you already have, and the referral ask you have been meaning to make since June. Neither one helps this quarter. Both of them decide the next one.

Protect the block. Put it on the calendar as an appointment, not an intention, and hold it through the close. The quarter after this one is being built right now, by whoever bothered.

The spike and the system.

Thirty-plus years carrying a number, a lot of those years now spent coaching managers rather than reps, and this is the pattern I would put money on.

A quarter you rescue in the last three weeks is a spike. It posts. It gets applauded at the sales meeting. And it teaches your organization exactly one lesson, which is that the last three weeks are when the real work happens. They will believe you, because you showed them. They will act on it. And you will spend the rest of your career running a business that only works when it is frightened.

The system is duller and it wins. A commit list that freezes. Three meetings that each ask one question. A strike count nobody argues with. A prospecting block that survives contact with a bad quarter. None of it is dramatic and all of it is repeatable, which is the entire point.

Isolate, Refine, Compound. See what you do. Change one thing. Let the years multiply it.

A quarter closes the way it ran. The only way to change how it closes is to change how it runs, and week eleven is far too late to start.

Run this.

The Week Ten Freeze. One page, four steps, run it on the Monday of week ten. It takes about forty minutes and it buys back three weeks.

1. Step 1. Print the commit list and date it. Every deal, the rep, the size as a percentage of the quarter’s number, and the customer-confirmed decision date. Not the rep’s date. The customer’s. If nobody can name the person on the buyer’s side who gave you that date, write “none” and keep going. The “none” column is your real risk list and it is always longer than the room expects. Any single deal over 40 percent of the number gets a circle around it and a second load-bearing deal assigned behind it.
2. Step 2. Draw the line, out loud. Say it in the meeting, in these words: “As of today the commit list is closed. Nothing new goes in. If something closes that was not on this page, we will celebrate it as a bonus, not count on it as a plan.” Then hold it. The first time you grant an exception the freeze is over and every person in that room will know it before you do.
3. Step 3. Run the strike count. Any deal committed three weeks running with no customer-side date change moves to next quarter. Move it in front of the team, not privately, and use the identical sentence every time: “Third week. It moves. If it closes early we will all be delighted.” Saying it the same way every time is what makes it a policy instead of a verdict on the rep. It also means the rep whose deal moves is not being singled out, which is the only reason anybody will ever tell you the truth about a date again.
4. Step 4. Reassign the hours you just freed. Take the time the team stopped spending re-arguing dead commits and put it somewhere with a yield. Two expansion conversations inside existing accounts and one referral ask, per rep, per week, held through the close. Expansion closes about 3 in 4 and referral about 1 in 5. Cold outbound, which is what everybody panics into in the first week of a quarter when the pipeline is empty, is closer to 1 in 100.

What it changes.

1. Time. Roughly a third of the forecast call disappears, because a deal with nothing new takes ninety seconds instead of ten minutes.
2. Opportunity. The next quarter opens with three weeks of built pipeline instead of three weeks of debt. That gap is the whole difference between a fast start and a slow one, and it has nothing to do with motivation.
3. Compensation. Reps stop carrying a number that was never real, which is the most quietly demoralizing thing you can hand a good seller. Their plan starts matching their calendar again.

The short version.

1. In the last three weeks of a quarter, a good team should be doing what it was doing in week two. A quarter you have to save was decided in week four.
2. Freeze the commit list in week ten. Nothing new enters after that. You lose the illusion of deals, not the deals.
3. Give each meeting one job. Pipeline review inspects next quarter. Forecast call inspects change only. One-on-one inspects behavior, not the number.
4. Third recommit and the deal moves out, said the same way every time so it is a policy and not a verdict.
5. Watch single-deal exposure hardest here. When one deal carries more than 40 percent of the quarter it fails to close about 4 times out of 5, and it is the deal nobody wants to be the one to move.
6. Protect the prospecting block through the close. Expansion closes about 3 in 4, referral about 1 in 5, cold outbound closer to 1 in 100. The close weeks are when the highest-yield entry points get abandoned.
7. The output of a leader is the output of the organization. Three weeks of a leader personally working deals is three weeks of one person’s output.

Here is your homework: put the Week Ten Freeze on your calendar now, before this quarter gets anywhere near week ten, and print the commit list the Monday it lands. Mark every deal that has a customer-confirmed date and every deal that does not, then circle anything carrying more than 40 percent of the number. Say the line out loud in that meeting: as of today the commit list is closed. The first exception you grant is the day the freeze stops existing, so decide before you walk in that there will not be one.

This essay ran as edition 15 of The Frontline Sales Forecast, the weekly newsletter. One issue a week, published here on the Friday, then emailed and posted to LinkedIn the same morning. Same piece everywhere. Get the next one.