A company we started working with last year was closing its books on a 19-business-day cycle, which meant the CEO didn’t have real August numbers until nearly the third week of September. Nobody had decided that pace on purpose.
It had just accumulated, one department at a time, until three weeks felt normal because nobody could remember it being any other way.
The instinct when a close runs long is to blame the tools. Get a better ERP, automate the reconciliations, buy something that promises to cut the close in half. Sometimes that helps. More often, the software isn’t the bottleneck at all. The bottleneck is a set of habits: waiting for information that could have been collected earlier, doing steps in sequence that could run in parallel, and treating every month as a fresh problem instead of a repeatable process with the same steps in the same order every time.
Start with what actually eats the first week of a slow close, because it’s rarely the accounting itself. It’s waiting. Waiting on the sales team to confirm which deals closed before month-end. Waiting on a department head to approve their credit card statement. Waiting on a vendor invoice that technically arrived on the 28th but didn’t get entered into the system until the 6th of the following month. None of this is bookkeeping work. It’s a series of small delays stacked on top of each other, and each one pushes the whole close back by however long that person took to respond.
The fix isn’t a faster accountant. It’s moving the deadline earlier and enforcing it as a real cutoff instead of a suggestion. A company that wants a five-day close typically needs departmental spend submitted and approved by the second business day of the new month, not the eighth. That means expense reports, corporate card coding, and any invoice that landed in the last week of the prior month all have a hard date attached to them, and that date gets treated the same way a payroll deadline gets treated: non-negotiable, because everything downstream depends on it.
Accruals are where a lot of the real time savings come from, and they’re also the part most teams do backwards. The traditional approach waits until month-end to figure out what wasn’t invoiced yet and then scrambles to estimate it. A five-day close builds standing accrual templates for anything recurring, things like utilities, rent, and known subscription costs, so the estimate is already built and just needs updating with current numbers rather than constructed from scratch every month. One controller we worked with kept a running list of the company’s fifteen most common accrual items with the prior month’s amount pre-loaded, and updating all fifteen took her under twenty minutes instead of the half day it used to take when she rebuilt the list from memory each cycle.
Subledger cutoff discipline matters more than most finance teams realize until they’ve fixed it. If accounts payable, accounts receivable, and payroll aren’t each closed and reconciled on their own before anyone touches the general ledger, the GL close becomes a moving target, because numbers keep shifting underneath it while someone’s trying to finalize the top-level financials. The teams that close fast treat each subledger as its own deadline with its own owner, closed independently by day two or three, so that by the time someone sits down to close the GL, nothing feeding into it is still in flux.
Bank reconciliations are the other place three weeks tends to hide. A monthly reconciliation done once, after the close, means someone is untangling thirty days of transactions all at once, and any discrepancy found on day 25 sends them backward through the whole month looking for where it broke. Reconciling weekly, or even daily for companies with meaningful transaction volume, means a discrepancy gets caught within a few days of when it happened, while the context is still fresh and the fix takes minutes instead of an afternoon.
Standardized journal entries solve a smaller but real problem: recreating the same entries from memory every month, slightly differently each time, because nobody wrote down the template. Depreciation, prepaid amortization, intercompany allocations, anything that follows the same formula every period should exist as a saved template that gets populated with current numbers rather than rebuilt, and it should be assigned to the same person every month so there’s no ramp-up time spent remembering how it worked last time.
None of this works without a close calendar that names a specific person and a specific day for every task, distributed to the whole team before the month even ends. Not a general sense of “we usually finish by the 15th,” but day one is subledgers, day two is accruals and reconciliations, day three is the GL close and preliminary review, day four is management review and adjustments, day five is finalized financials out the door. When everyone knows exactly what’s due and when, most of what used to happen in sequence, one person waiting for another to finish before they can start, ends up happening in parallel instead.
The company on the 19-day cycle got to a five-day close in about two months, without changing accounting software once. What changed was the deadline structure, the accrual templates, and a calendar that made the sequence of the close visible to everyone doing it. The CEO had July numbers by the first week of August. Nobody on the team would go back.