How to Run an End of Quarter Close Process

What is an end of quarter close process?

A published set of rules for the last three weeks of the quarter that removes improvisation from the highest pressure period you have.

Without one, the final weeks run on escalation. Discounts get approved over instant message, deal desk becomes a bottleneck at the worst possible moment, and legal receives four contracts on the final afternoon. The revenue may still land, but nobody can explain afterward what it cost.

The close process is a schedule and an approval structure. It says what happens each week, who decides what, and which actions require a written record.

When does it start?

Week 10 of a 13 week quarter, and the timing is not arbitrary.

The third month of a quarter usually closes more than the first two, and Q2 and Q4 usually run stronger than Q1 and Q3. That shape is predictable enough to staff against. Deal desk capacity, legal review, and finance availability should be scaled up before the volume arrives rather than in response to it.

Teams that treat quarter end as a surprise every quarter are choosing to be surprised. The pattern shows in your own historical data, and the fix is a calendar.

What happens each week?

Assign each week one job and one output. Week 10 does the most work. A deal that has no signature path identified with three weeks left is not closing this quarter, and naming that early frees the team to work deals that can.

What should be off limits?

Improvised discounting and unlogged pull-forward.

Both are legitimate levers. Neither should be pulled without a record. Pull-forward is the more expensive of the two because its cost lands in a period nobody is looking at yet. Teams consistently understate what it takes out of the next quarter, then start that quarter behind and repeat the maneuver at a larger scale.

Set two rules and hold them:

- Any discount beyond the standard band requires deal desk approval with the reason recorded on the opportunity. - Any deal pulled forward from a future period is logged with the amount, the original close date, and the resulting hole in that period.

Then report the totals in the quarter retro. When leadership sees how much of the quarter came from next quarter's pipeline and at what additional discount, the conversation about how the number was made becomes as serious as the conversation about whether it was made.

How do you handle deals that slip?

A commit deal that slips past quarter end leaves commit and is re-qualified before it can go back.

A rep changing a close date is the strongest single signal that a deal is in trouble, and a deal that slips across a quarter boundary is less likely to close even when it stays in commit. Carrying those deals forward at full confidence is how a new quarter starts with an inflated number and a stale plan. Track this deliberately, because repeated deal slippage on the same opportunities is the clearest pattern in most pipelines.

Re-qualification should be short and specific. Confirm the economic buyer is still in place, the budget still exists in the new period, and there is a dated next event. Deals that pass return to a forecast category. Deals that do not go back to an earlier stage where they belong.

Watch for the quieter version too. A deal with no change in stage, close date, or amount for weeks is not waiting, it is over. Across ORM's customers it varies, but more than 10 percent of pipeline typically sits untouched for a year, and quarter end is when that inventory gets quietly recycled into next quarter's coverage.

What happens the day after the quarter ends?

Freeze the actuals and compare them against what you believed at day 1, day 30, and day 60.

This is the step that converts a close into a learning cycle. Careful manual forecasting of new and expansion business usually lands near 90 percent accuracy, and it costs real effort and goes stale between refreshes. The interesting question is not the final accuracy number, it is how early the forecast was right.

Getting the forecast right in the final week of the quarter does not help anyone, because by then the quarter has already happened. The value sits in knowing the likely shape of the quarter on day one, early enough to change it. Measuring forecast accuracy at day 1 and day 30 rather than only at the finish line is what makes that improvement visible.

Run the retro within five business days while the reasons are still remembered. Four questions cover it: which committed deals did not close and why, which unexpected deals did close, how much of the number came from pull-forward or discount, and what changed in the market that the model did not catch. That last question is the one that matters most. Forecasts miss because something changed in the business or the market and the model was still running on old assumptions, so the retro is where you find the change and update the assumptions before the next quarter repeats the miss. If you are rebuilding the process itself, start from how to forecast revenue.