The short answer: Most month-end close errors are timing errors, not classification errors. They reconcile perfectly, net to zero across the year, and still make every monthly margin number wrong. One $61,000 vendor invoice booked a month late moves 7.2 points of gross margin out of one month and into the next, while the two-month average stays exactly right.
- A missed expense cutoff overstates one month’s gross margin and understates the next by the identical amount. The annual P&L never reveals it.
- Standing accruals that are never re-based drift. An estimate running $8,300/mo under actual dumps $24,900 into the true-up month, or 2.9 points of margin.
- An allocation driver set 14 months ago and never refreshed misstates one product line by 8.4 points while total company margin stays correct.
- Review thresholds measured against revenue hide errors that are enormous against income. The same $61,000 is 7.2% of monthly revenue and 72% of an $84,500 monthly pretax result.
Last updated September 2026.
A reconciliation proves a balance. It does not prove a period. Books can tie every bank account, agree every subledger to the general ledger, and still report a gross margin off by seven points. The error sits in which month a cost landed, not in whether it was recorded.
That is why these errors survive. They pass the checks a close is designed to run. The four below are the ones Debit & Co. finds most often after inheriting a set of books that already closes on time.
Where do month-end closes actually go wrong?
In period assignment. The account is right and the month is wrong. Four failures produce most distorted margins: expense cutoff, accruals nobody re-bases, allocation drivers left stale, and review thresholds pointed at the wrong denominator.
Classification errors are a different problem. A cost sitting in the wrong account distorts margin permanently, and the bookkeeping mistakes that distort the P&L cover that ground. Everything here assumes the coding is correct.
| Close error | Why it still reconciles | Effect on reported margin | Where it surfaces |
|---|---|---|---|
| Expense cutoff missed | Bank and AP subledger both tie; the invoice is recorded, just later | One month overstated, the next understated, by equal amounts | Gross margin sawtooth between adjacent months |
| Standing accrual never trued up | The accrual account has a balance and a schedule behind it | Two or three months flattered, then one month absorbs the catch-up | A single outlier month with no volume explanation |
| Stale allocation driver | Total expense is correct; only the split is wrong | Company margin correct, product or segment margin wrong | Product P&Ls that contradict the sales team’s experience |
| Threshold on the wrong denominator | The variance clears a revenue-based materiality test | Errors worth most of a month’s income never get reviewed | Nowhere, until an auditor or a buyer looks |
How does a missed expense cutoff move gross margin between two months?
It relocates margin from the later month to the earlier one, or the reverse, without changing the total. The governing rule is narrower than it usually gets stated. Costs arising directly from the same transaction as the sale, cost of revenue above all, belong in the period of the related revenue. ASC 330-10-10-1 makes that the stated objective of inventory accounting.
Many operating costs work differently. Administrative salaries are recognized as incurred, and depreciation is allocated on a systematic basis, neither of them tied to a particular sale. Gross margin sits entirely in the first category. That is why cutoff errors damage it more than they damage operating expense.
Work the numbers. A services business reports $845,000 in monthly revenue and $270,400 in cost of revenue, a 68.0% gross margin. A contractor performs $61,000 of July delivery work and invoices on August 12. Nobody accrues it in July.
- July as reported: $845,000 − $270,400 = $574,600, a 68.0% gross margin.
- July as it should read: cost of revenue rises to $331,400, leaving $513,600, a 60.8% gross margin.
- August as reported: the $61,000 lands there, so August shows $331,400 of cost and the same 60.8%.
- The two months blended: $1,690,000 of revenue against $601,800 of cost, 64.4% either way.
The swing is 7.2 points, and it runs in both directions at once. July is flattered by exactly what August is penalized by. Any decision made off July’s 68.0% was made off a month that actually delivered 60.8%.
The tax rules describe the same discipline from a different angle. IRS Publication 538 takes an expense into account under the accrual method once the all-events test is met and economic performance has occurred. Liability fixed, amount reasonably determinable, service actually rendered. All three were true in July.
Why do standing accruals nobody re-bases distort a full quarter?
Because the estimate is right when it is set and wrong every month after. A recurring accrual is a placeholder for an amount not yet known. It earns its place only if somebody compares it to actual invoices and re-bases it.
Take a cloud hosting accrual entered at $42,300/mo. Usage grows and invoices settle at $50,600. The gap is $8,300/mo, small enough to clear any variance report. Nothing flags for three months.
Then the true-up arrives. Three months of shortfall, $24,900, posts in a single period. On $845,000 of revenue that is 2.9 points of gross margin removed from one month and quietly borrowed from three. The distortion is worse than the cutoff case in one respect: it does not reverse cleanly, so the pattern reads as a business event rather than an error.
The control is a re-basing cadence, not a bigger spreadsheet. Compare every standing accrual to trailing actuals each quarter. Re-base anything off by more than 10%. Retire any accrual whose underlying invoice now arrives before the books close.
What does a stale allocation driver do to product-level margin?
It leaves total company margin intact and makes each product line wrong. That combination is dangerous, because the number executives trust most is the one that still looks right.
A company splits $118,000/mo of shared infrastructure between two product lines. The 60/40 split was set 14 months ago from a usage sample. Consumption has since shifted to 38/62.
The arithmetic is direct. The split is off by 22 points, so $118,000 × 0.22 = $25,960 of cost sits against the wrong product every month. Product A carries $310,000 in monthly revenue, which makes that misallocation 8.4 points of its gross margin. Product A looks worse than it is. Product B looks better, and the roadmap follows the flattered line.
Drivers decay because they are stored as constants. Refresh them each quarter from the systems that measure the actual consumption: seat counts, compute hours, ticket volume, headcount by function. A driver that has not moved in four quarters is usually a driver nobody has checked.
Which month-end close errors slip under the materiality threshold?
The ones measured against the largest number on the page. Most review thresholds are set as a percentage of revenue, which is the denominator most forgiving of a period error.
Return to the $61,000 invoice. Against $845,000 of monthly revenue it is 7.2%. Against an $84,500 monthly pretax result it is 72%. One threshold sends it to a controller, and the other lets it through. The error never changed.
Regulators reached this conclusion long ago. SEC Staff Accounting Bulletin No. 99 rejects exclusive reliance on any single quantitative benchmark, including the familiar 5% rule of thumb, and directs preparers to weigh qualitative factors alongside size. A misstatement that flips a segment from growth to decline is material at any percentage.
Private companies inherit the logic even without the filing obligation. Set the threshold against pretax income or against gross margin points, whichever bites sooner. Then apply a qualitative override: any error that changes a margin trend direction, a covenant calculation, or a product-level decision goes to review regardless of size.
How do you catch period errors before the close locks?
Test for the pattern rather than the balance. Period errors have a signature that reconciliations cannot see, and four checks expose nearly all of them.
- Run the sawtooth test. Chart gross margin by month for 13 months. Two adjacent months deviating in opposite directions by more than 2 points, with no volume or pricing explanation, is a cutoff error rather than a business event.
- Hold accounts payable open through business day 3. Sweep vendor commitments, signed statements of work, and contractor hours before cutoff instead of waiting for invoices to arrive.
- Re-base every standing accrual quarterly. Any estimate more than 10% from trailing actuals gets a new number and a documented reason.
- Keep a period-error register. Log each timing correction with the month it belonged to. Repeat entries against the same vendor identify a broken intake process, not a careless month.
Tooling helps at the detection layer and stops there. An AI-native ledger such as Puzzle gives you the data; a team turns it into a close. Deciding which period a half-finished engagement belongs to is a judgment. The Continuous Close Method™ puts that judgment on a schedule, catching cutoff items during the month.
Aaron Ressel reviews the 13-month margin trend on every client packet before it ships, ahead of the balance-sheet reconciliations. The pattern is faster to read than the ledger. For the sequencing that makes these checks fit inside a working calendar, see the month-end close process and the discipline of variance analysis before the close locks.
Frequently asked questions
What are the most common month-end close errors?
Timing errors dominate. Expenses land in the month the invoice arrived rather than the month the work occurred. Standing accruals go uncompared to actual invoices, and allocation drivers sit at values set several quarters ago. Each one reconciles cleanly, which is why routine close checks miss them.
How does a cutoff error affect gross margin?
It moves margin between two adjacent months without changing the total. A $61,000 cost booked one month late overstates the first month’s gross margin by 7.2 points on $845,000 of revenue and understates the second by the same amount. The two-month blended figure of 64.4% stays correct throughout.
Why do my monthly margins swing when revenue is stable?
Stable revenue with volatile margin usually indicates a period error rather than an operating change. Look first for adjacent months that deviate in opposite directions, then for a single outlier month absorbing an accrual true-up. Both patterns are visible on a 13-month gross margin chart.
What materiality threshold should a private company use for close errors?
Set it against pretax income or gross margin points rather than revenue. Add a qualitative override for anything that changes a trend direction, a covenant calculation, or a product decision. SEC Staff Accounting Bulletin No. 99 rejects exclusive reliance on a single percentage, and the reasoning applies to private books.
Does a stale cost allocation change total company margin?
No, and that is the difficulty. Total expense is correct, so consolidated gross margin is right while every product or segment line is wrong. A $118,000 monthly pool split 60/40 when actual consumption runs 38/62 misplaces $25,960 each month, or 8.4 points of margin on a $310,000 product line.
How do you prevent period errors in the month-end close?
Hold accounts payable open through business day 3 and sweep vendor commitments before cutoff. Re-base every standing accrual quarterly against trailing actuals, refresh allocation drivers from consumption data, and run a 13-month margin trend before the reconciliations. Log each timing correction so repeat sources become visible.
