The six-day close: where a finance team's month-end actually goes

Pull up a month-end close checklist. Not a vendor's marketing version, an actual one: the spreadsheet a controller keeps, forty to sixty line items long, each with an owner and a due day. Read it top to bottom and a pattern shows up fast. The first third is reconciliation. The middle is adjustments and accruals. The last third is consolidation and reporting. Only the reporting rows need a person reading the result and deciding what it means.

That shape is why the close takes as long as it does. APQC's benchmark of 2,300 organizations puts the median monthly close at 6.4 calendar days, with top performers at 4.8 or less and the slowest quarter needing ten days or more. Half of finance teams take six or more business days, and only 18 percent finish in one to three. The days are not going into judgment. They are going into matching, keying, and chasing.

Get your close mapped room by room and see which two days are pure reconciliation

Where the six days go

The close is not one task. It is four rooms the work moves through, and they are not equally hard. Map them before you buy anything.

The reconciliation room is where the hours pile up. The Ledge 2025 benchmark puts cash and balance-sheet reconciliation at 20 to 50 hours a month, with teams jumping between three and five systems to finish it. That is not a judgment problem. A rules engine matches deposits to the bank feed, flags the handful that do not tie, and hands a person only the exceptions. The same holds for the adjustments room: an accrual that repeats every month with a predictable formula is a template, not a decision.

What month-end close automation actually changes

Month-end close automation replaces the mechanical two-thirds of the close with rules and leaves the judgment two-thirds of the value with your team. In practice that means three moves. Reconciliations run continuously against live feeds, so the first business day of the close starts with matched accounts instead of a blank sheet. Recurring journal entries and accruals post from templates on a schedule, so no one rekeys the same entry twelve times a year. Variance and flux commentary gets drafted from the ledger movements, so the controller edits a first pass rather than writing from scratch. What does not change: the estimates, the unusual transactions, the reserve calls, and the final sign-off. Those stay with a licensed, accountable person. Automation clears the mechanical rows so the team can spend its days on the estimates and the reading of the result, which is the work that pays.

The friction is the spreadsheet

Ask finance teams why the close drags and they do not blame the accounting. They blame the plumbing. In the Ledge survey, 94 percent still run the close in Excel, and half of them name Excel itself as a reason it is slow. The rest of the root-cause list reads the same way: 56 percent point to waiting on other departments, 50 percent to Excel-driven process, 40 percent to legacy systems that do not talk to each other, and 37 percent to being short-staffed. None of those is a knowledge gap. They are handoffs and copy-paste, and both are exactly what software is good at removing.

This is the part operators miss. The close is slow not because the team lacks skill but because the work is stitched together by hand across systems that were never connected. Fix the stitching and the six days compress on their own.

The order we would run it

When we map a finance team's close, we do not start with the flashiest tool. We start where the volume and the rules are highest, because that is where the hours actually sit. The order:

1. Reconciliations, first and always. Highest volume, most rules-based, least judgment. This is the same first move we would make on a standalone bank rec, and it clears the biggest single block of hours.
2. Recurring journal entries and accruals. Anything that repeats with a formula becomes a scheduled template. The upstream feeders matter here too: a clean accounts payable run and a current receivables ledger mean fewer surprises at cutoff.
3. Variance and flux commentary. The system drafts the explanation from the numbers; the controller corrects and signs. This is assistive, not autonomous.
4. The close calendar itself. Owners, dependencies, and status in one place, so the 56 percent cross-team wait becomes visible instead of a series of chased emails.

Notice what is not on the list to automate: the estimates, the reserves, and the sign-off. Those are the reporting room, and they are why an accountant is in the building. A Chicago services firm closing three entities has the same first two moves as a single-entity manufacturer. The volume rooms come first because that is where the days are.

The trap: closing wrong faster

Here is the honest caveat, and it is the one that sinks most close projects. Automation does not fix bad data. It repeats it faster. If your chart of accounts is a mess, if the same expense lands in three different accounts depending on who codes it, if half your reconciliations are really plugs, then automating on top of that produces wrong statements sooner and with more confidence. The fastest close in the world is worthless if the numbers are off. So the first week is not a software install. It is a cleanup: fix the mapping, kill the plugs, and standardize the coding. Then automate. A slower correct close beats a fast wrong one every month of the year.

Go back to that checklist. Cross out every reconciliation row, every recurring accrual, every entry that posts the same way each month. On most teams that pen goes through more than half the page. What is left, the estimates and the reading of the result, is the close that actually needs a person and the six days you should be spending it on.