Why the close drags
A slow close is almost never about effort — it's about waiting. Waiting on a bank statement to post, a missing receipt from someone who is travelling, an unreconciled account nobody has looked at since spring, an approval sitting in an inbox. The bookkeeper is not slow; the bookkeeper is blocked. Speed comes from removing the waiting, not from working harder in the final 48 hours.
That distinction changes what you do about it. If the close is an effort problem, the fix is more hours or more people, and it costs money every month forever. If it's a waiting problem, the fix is upstream — earlier information, cleaner feeds, fewer unknowns arriving on day one of the close — and it's mostly a one-time cost. Nearly every SMB close that takes fifteen days is a waiting problem being treated as an effort problem.
The other reason the close matters is timeliness. Numbers that arrive on the twenty-second are a history lesson. Numbers that arrive on the fifth are a decision tool, and the difference between the two is the entire value of the exercise: a P&L you can act on while the quarter is still in play versus one you read after the outcome is fixed.
The close checklist
The order matters here, because each step depends on the ones before it. Recording accruals before cash is reconciled means recording them against numbers that are about to change.
- Reconcile every bank and credit-card account. No close is real until cash ties to the statement to the penny. This is the foundation — everything downstream inherits any error here. When it won't tie, work it systematically rather than by hunting: troubleshooting an out-of-balance reconciliation walks the usual culprits, and deposits in transit and outstanding checks explain the timing differences that are supposed to be there.
- Clear the suspense and clearing accounts. Undeposited funds and payment-processor clearing accounts should end the month at or near zero. A balance sitting in one means a deposit was recorded twice, or a payment was received and never matched. These are the errors that surface months later as a mystery, so catch them while the transactions are still recent enough to remember.
- Reconcile AR and AP subledgers to the general ledger control accounts. The A/R aging and A/P aging totals must equal the control account balances. When they diverge, something was posted directly to the control account instead of through an invoice or bill, and finding it now is far cheaper than finding it at year-end.
- Reconcile payroll liabilities. Withholdings and employer taxes you're holding should match what's actually owed and remitted. Reconciling payroll liability accounts is the step most often skipped, and it's holding other people's money — as is sales tax, which deserves the same treatment for the same reason.
- Record accruals. Expenses incurred but not yet billed, and revenue earned but not yet invoiced. This is what makes the month's P&L reflect the month's activity rather than the month's mail — see accrued expenses for the mechanics, and note that this step is what separates an accrual close from a cash-basis one.
- Record prepaids and depreciation. Amortize prepaid expenses for the portion consumed this month and run the fixed-asset depreciation schedule. Both are pure schedule work, which makes them ideal candidates for templates.
- Post the remaining adjusting entries. Deferred revenue recognized this period, inventory adjustments, bad-debt provisions, anything else that moves the books from what happened in the bank to what happened in the business. Adjusting entries covers the standard set.
- Review the P&L and balance sheet for anomalies. Compare to budget and to prior month; investigate any line that moved more than your materiality threshold. A comparative P&L makes this a two-minute scan instead of a hunt. Then review the balance sheet — every balance should be one you can explain and support with a schedule.
- Lock the period. Once reviewed, close the period so no one back-dates an entry into a month you have already reported on. This is the step that makes the close mean something; without it, last month's numbers are still capable of changing after you've shown them to someone.
The twelfth month-end of the year carries one extra step the other eleven don't: the year-end closing entries that zero out revenue and expense accounts into retained earnings and reset the books for a clean new fiscal year.
Close a little every day
The biggest single lever is continuous reconciliation. If you reconcile cash weekly and code bank-feed transactions daily, month-end is a review rather than a reconstruction. Teams that save it all for the close guarantee a slow close, because they've concentrated a month of work into the one week when it's most urgent and least pleasant.
The compounding effect matters more than the arithmetic. A transaction categorized the day it lands is categorized by someone who remembers what it was; the same transaction categorized three weeks later is a research project involving a vendor's website and someone's calendar. The cost per transaction is not constant — it rises the longer you wait, which is why a distributed close is genuinely cheaper and not merely differently scheduled.
Bank-feed rules do most of this work automatically for recurring vendors once you've set them, which turns daily coding into a short review of exceptions rather than a data-entry shift.
Standardize and document
- A written close calendar with owners and due dates for each task. The point is less the schedule than the ownership — most close delays are tasks nobody thought were theirs.
- Reusable journal-entry templates for recurring accruals, so the monthly accrual is a review of a number rather than a fresh derivation.
- A binder of supporting schedules — depreciation, prepaids, accruals, deferred revenue — updated each month. This is also most of what an auditor or lender will ask for, so maintaining it monthly is audit-readiness as a byproduct rather than a project.
- A stable chart of accounts. Restructuring accounts mid-year is one of the reliable ways to make comparative reports meaningless and the close slower.
What "closed" should actually mean
Be honest about the standard, because a close that isn't locked isn't closed. A period is genuinely closed when every balance-sheet account has a supporting schedule that explains it, the period is locked against back-dated entries, and the reports you produced from it will still say the same thing if you run them again in six months.
That last condition is the one that gets violated most often and noticed least. If last month's P&L can still change because someone posted a late entry into it, then every report you've circulated is provisional and nobody downstream knows which version they're holding. Locking the period is what converts your numbers from a running estimate into a statement of record.
Measure your close
Track days to close every month and treat it as a real operating metric. A five-day close is excellent for an SMB; ten is normal; fifteen-plus means the process, not the people, needs fixing.
Track where the days go, too, not just how many there are. Most closes have one or two dominant blockers — a credit-card statement that posts late, one vendor who invoices on the tenth, an approval that always waits on the same person — and the fix is usually specific and small once it's named. A close that improved from fifteen days to nine almost never did it by getting uniformly faster; it did it by eliminating two blockers.
Hosting Books supports this with connected bank feeds and reconciliation that run continuously rather than only at period end, an audit log that shows what was posted and when, and period locking so a closed month stays closed. The checklist above is still yours to run — the software removes the waiting, not the judgment.
This article is general educational information about accounting concepts and is not accounting or tax advice for your specific situation. Close requirements, materiality thresholds, and reporting obligations vary — consult a qualified professional for your circumstances.