The monthly close checklist: the 10 steps, in the order they have to run
Updated: Aug 20
Written By: Hemanth Aligingi, Senior Executive at Outsource360
On 13 February 2024, just after the closing bell, a US-listed ride-hailing company published its fourth-quarter earnings release. Every monthly close checklist ends with the same step, and it's the step this company skipped. One line in that release stated that adjusted margin would expand by 500 basis points over the coming year.
The correct figure was 50. One extra zero, sitting in a document that had already passed through a finance department, an external audit relationship and a formal disclosure process. Shares surged in after-hours trading, and the move was reported across outlets at roughly 60 to 67 percent.
Then it got stranger. The company's finance chief corrected the figure live on the analyst call that followed, and the chief executive took responsibility the following day, describing it as a zero that slipped in.
Even after the correction, the stock closed roughly 35 percent higher on the strength of the underlying results. That's the uncomfortable part. The number was wrong for hours, it moved billions in market value, and nothing inside the process caught it before the market did.
So what actually failed? Not capability. That company had a finance department, external auditors and a disclosure process most founders would envy. What it didn't have on that particular evening was somebody reading the finished numbers as an outsider, before they went out.
That last step has a name in every accounting textbook and almost no presence in a founder's month. The books get produced: transactions categorised, accounts reconciled, statements generated. Then nobody independently reads them, because the person who would read them is the person who made them.
It's a structural gap, not a discipline problem. You can be careful, fast and genuinely well organised and still have no reviewer, for the simple reason that there is only one of you.
Most month-end guides are written for a finance team with a reviewer built into the org chart: a bookkeeper posts, a senior checks, a manager approves and signs off. A founder closing their own books has every one of those steps except the last, and the last is the one that catches an extra zero. That's the gap this guide is built around.
So the checklist below is ordered the way it is for a reason. It ends with a test rather than a feeling, it maps the outside events that actually gate your calendar, and it carries a full section on what to do about the reviewer you don't have. Run it cold the first time and it eats an afternoon. Run it every month and it settles into about an hour.
First, the definition, because a surprising number of founders are running something they'd call a close and a bookkeeper wouldn't.
A monthly close checklist is the fixed order of steps that turns a month of raw transactions into books you can trust: record, reconcile, adjust, agree, report, lock. Guides commonly quote five to seven working days. In practice the date is set by when your last external input arrives.
That's the shape of it. What follows is the running order in full, the day-by-day calendar that tells you what you're actually waiting for, the mechanics of locking a period, and the six-condition test that proves a month is genuinely closed.
On this page
The acceptance test: six conditions a closed month has to pass
Closing alone: controls when you are the entire finance function
What the monthly close is, and what it is not
Plenty of founders think they're closing the books when they're really catching up on data entry. One produces numbers you can act on, the other a tidier spreadsheet. So what separates them?
The close in one sentence, and the four things it produces
The monthly close is the fixed sequence that turns a month of raw transactions into finished, locked accounts: record, reconcile, adjust, agree, report, lock. Six verbs, always in that order. Out come four things: a trial balance that balances, a profit and loss statement, a balance sheet and a cash flow statement, each tied back to evidence from outside your own ledger.
Where the daily work ends and the close begins
Bookkeeping is the daily habit: coding transactions, raising invoices, filing receipts. The month-end close process is the periodic control that checks the daily habit was right. Finance people draw one more line, between the accounting close (the ledger work described here) and the financial close, which adds consolidation, disclosure and formal sign-off for entities that report externally. In practice, a founder running one entity is doing an accounting close, and that's the correct scope.
Monthly, quarterly and annual: three closes, one ledger
Here's the good news. You run one process, and the quarterly and annual versions bolt extra work on top.
Close | What triggers it | What it adds on top of the monthly close | Who reads the output |
Monthly | The calendar month ends and your last external input arrives | Nothing. This is the base procedure every other close is built from | You, plus anyone you report to internally |
Quarterly | Quarter end, investor reporting cycles, and indirect-tax or payroll filing periods in markets that run them quarterly | Aggregation of three closed months, quarter-on-quarter variance review, and any quarterly return the entity owes | Investors, lenders, board members, and tax authorities where a quarterly return is due |
Annual | The financial year end | Fixed-asset and depreciation schedules, stock counts, accrual true-ups, statutory accounts and the tax computation | Your accountant, the company registry, and the tax authority |
A founder with a Delaware C-corp and an Indian delivery arm still keeps one ledger per entity and closes each monthly. But the pitfall is treating the close as a catch-up window. If the only bookkeeping you do all month happens in the first week of the next one, you aren't closing. You're reconstructing, and reconstruction is where errors breed.
Before you start: the four things that have to be true
For a small business closing the books without a finance team, preparation does most of the work. Four conditions decide whether the close takes an hour or swallows a weekend, and all four are set before the month ends (before, not during).
A business bank account that is only ever the business
Mixed personal and business spending is the most expensive habit in founder bookkeeping, because it turns a reconciliation into an investigation. We'd recommend one account, one card, one processor, all in the company's name. If a personal transaction slips through, code it to a director's or owner's account the day it appears (not at close).
A chart of accounts you did not change mid-month
This is where most founders go wrong: renaming or merging accounts partway through a period makes your comparatives lie. Freeze the chart on the first of the month and apply any restructuring from day one of the next. Thirty days of tidiness buys back a fortnight chasing variances that were never real.
A cut-off you actually hold
The last day of the month isn't an administrative formality. It's the line that decides which period a transaction belongs to, whether or not the money has moved. So what do you do on the day? Submit expenses, raise every invoice you've earned, then stop posting.
One named place where the evidence lives
And every reconciliation needs a document behind it: the bank statement, the payroll report, the processor's fee summary. Decide once where those live (a dated folder per month is enough) and file them as they arrive. The pitfall is timing. Start reconciling while the feed is still importing the last two days of the month and you'll reconcile everything twice.
The monthly close checklist, in the order it has to run
Close the cut-off and stop posting. Fix the last day of the month as the line, and post nothing new into the period once you begin.
Import and code every transaction. Pull in every bank, card and processor feed until the transaction count stops moving.
Chase and enter unrecorded bills and invoices. Anything you owe or are owed for work delivered in the month belongs in the month.
Reconcile every bank and card account. Agree each account to the statement the bank issued, not to the feed the software imported.
Reconcile payment processors to gross, fees and payouts. Split the processor's activity into gross charges, platform fees, refunds and chargebacks before you touch revenue.
Post accruals, prepayments and depreciation. Move costs and income into the period they belong to rather than the period they were paid in.
Clear the suspense account to zero. Every item parked there gets a home before the month can close.
Agree sub-ledgers and control accounts to the ledger. The payables and receivables sub-ledgers, and every control account, must tie back to the general ledger.
Produce and read the four statements. Run the trial balance, profit and loss, balance sheet and cash flow statement, then actually read them.
Lock the period and file the evidence. Set the lock date and put every supporting document in the month's folder.
That's the whole procedure. Nothing is gated, nothing sits behind an email form, and the order is not negotiable.
Steps 1 to 3: everything in, before anything is checked
The first three monthly close steps share one job: completeness. You cannot check books that are still missing transactions. But chasing unrecorded items is where founders lose a day. A supplier invoice dated the 6th for work done on the 22nd belongs in the earlier month (and this is the part most guides skip).
Steps 4 and 5: reconcile to an external statement, not to your own records
The short answer: reconciliation only means something when the reference point comes from outside your ledger. Tick your bank account against the statement the bank issued, closing balance included, not against the imported feed (feeds drop, duplicate and re-date transactions more often than founders expect). But the same logic covers cards, loans and every processor.
Steps 6 and 7: the adjusting entries a founder genuinely needs
Accruals put a cost into the month it was incurred even though the invoice hasn't arrived. Prepayments do the reverse, spreading something already paid across the months it covers: annual software licences and insurance are the usual candidates. And depreciation writes down fixed assets on a schedule you set once and repeat. The suspense account holds transactions you couldn't code at the time, and it must read zero before the month closes.
Step 8: sub-ledger and control-account agreement
Agree the closing balances on your payables and receivables sub-ledgers to their control accounts in the general ledger, then review both ageing reports. If you want the full picture of how the payables and receivables ledgers differ in practice, that's a separate subject; here the only job is agreement.
The same goes for sales tax or VAT, payroll liabilities and any owner's or director's account. If you're GST-registered in India, this is where you check that the input tax credit in your books agrees with the auto-generated statement before you file, and that deducted tax sits in the correct section-wise ledger. Section 16(2)(aa) of the CGST Act makes that first check a condition of the claim.
Steps 9 and 10: produce, read, lock
Producing the statements takes one click. This is where most founders go wrong. Reading them is the step with value in it, and most glance at the bottom line and move on.
Compare each line to the prior month and to your expectation, then write one sentence explaining anything that moved materially. Then lock, because an unlocked period stays editable.
The steps you can skip at your size, and the two you cannot
Do you need to count inventory every month? Only if you hold stock and its value is material. Petty cash is the same conditional: on negligible amounts, count quarterly and write the policy down rather than claiming a count you don't run.
But two steps have no skip clause. Reconciling every bank, card and processor account to an external statement is one, and clearing suspense to zero is the other. Skip either and the numbers you produce are decorative.
Why the order matters, and what breaks when you invert it
A month-end close checklist isn't a to-do list with numbers on it. The sequence encodes dependencies, and inverting two steps can quietly cost you a day of rework. So what actually breaks?
Reconcile before you adjust, never after
Here's the thing. Adjusting entries assume the underlying transactions are complete and correct. Post accruals and depreciation first, then reconcile, and every difference you find forces you to work out whether it belongs to the raw data or to your own journal. That's why a checklist makes a measurable difference: it removes the decision about what to do next at the moment you're tired.
The suspense account is a queue, not a bin
Suspense is temporary storage for transactions you couldn't identify at the time. Used as a queue it's healthy. Used as a bin it becomes the place uncomfortable items go to be forgotten, and a balance sitting there at year end is one of the first things an accountant will query. Clear it monthly, and if an item genuinely can't be identified, code it somewhere defensible and note why.
Speed and accuracy were traded against each other from the start
The "fast close" isn't a natural law. It's a compliance artefact of the mid-2000s, when Sarbanes-Oxley and the SEC's accelerated-filer regime cut the annual filing deadline from 90 days to 75, and to 60 for the largest filers, and the quarterly deadline from 45 days to 40. Accounting research on the 2003 acceleration found that restatement likelihood actually rose for firms forced to file faster. Worth remembering when a vendor sells you a three-day close.
What a review-first close means when the review is yours
Based on what we've seen, a review-first close builds the checking in rather than bolting it on at the end. Each block finishes with its own verification (feeds complete, statements agreed, suspense at zero) instead of one big review that never happens. And when you're the only reviewer, that distributed checking is most of what you've got.
The close calendar: what gates each day
Almost every guide answers "how long should a monthly close take?" with a number and stops there. The number is useless on its own, because a close isn't gated by how fast you work. It's gated by when the outside world sends you things. So what is your month end close timeline actually waiting for?
Day 0 to Day 7, and what lands when
Day | What lands | What you can do | What stays blocked |
Day 0 (last day of the month) | Nothing external. Your own cut-off is the only event | Submit expenses, raise every invoice you've earned, stop posting into the period | Every reconciliation, because the month isn't finished yet |
Day 1 | Bank and card feeds finish importing the final days of the month | Code every transaction, chase unrecorded supplier bills and customer invoices | Bank reconciliation, because most statements haven't been issued |
Day 2 | Month-end bank and card statements start to issue | Reconcile each account whose statement has arrived, closing balance included | Accounts on a non-calendar statement cycle, and all processor work |
Day 3 | Payroll finalises for the month | Post the payroll journal and the payroll liability entries | Payroll control-account agreement until the final payroll report exists |
Day 4 | Payment-processor settlement completes and the monthly fee report publishes | Split gross charges, fees, refunds and chargebacks; clear the processor clearing account | Revenue. It isn't final until the processor split is posted |
Day 5 | Nothing external | Post accruals, prepayments and depreciation; clear suspense to zero | Nothing, provided Days 1 to 4 are genuinely complete |
Day 6 | Nothing external | Agree sub-ledgers and control accounts; produce and read the four statements | Sign-off, while any material variance is still unexplained |
Day 7 | In India, deducted tax falls due by the 7th of the following month | Lock the period, file the evidence, pay what's due | Nothing, once the period is locked |
The five external dependencies that actually set your timeline
Here's the thing. Five inputs arrive on somebody else's schedule, and between them they decide your date.
Bank and card statement availability comes first, because a reconciliation without a statement isn't a reconciliation. Payroll finalisation comes second, and it moves if your provider runs a cut-off you don't control. Third is payment-processor settlement plus the monthly fee report, which is usually the last thing to land.
The final two are India-specific and they're the hardest dates in this calendar. Deducted tax has to be paid by the 7th of the following month, section by section. And the auto-generated input-tax-credit statement becomes available on the 14th, which is what your books have to agree with before the monthly return goes in.
Your close is five days only if...
Let's be honest about the five-day close. It is achievable, but only under conditions worth stating plainly. Your bank statements issue on the 1st or 2nd, payroll is finalised in-month, and you take money through one processor rather than four.
Nobody is still emailing you receipts on the 9th. You hold no stock, and you have no filing obligation that depends on a portal publishing data mid-month.
Change any one of those and the five-day close becomes a seven or ten-day close, without anybody working less hard. On a finance-professional Q&A community, practitioners describing their own real closes reported anything from about two days at the fast end to roughly 45 to 60 days at some privately-held companies. That's a thirty-fold spread among competent people.
A long close is usually a dependency problem, not an effort problem
When a close runs long, founders reach for discipline first: earlier starts, longer evenings, a stricter checklist. Frankly, this gets overlooked: the fix is almost always upstream. Move the processor to a shorter settlement cycle, switch a supplier to a statement date that suits you, bring payroll cut-off forward by three days, or stop accepting receipts after the 3rd.
Is a long close a process problem or a staffing problem? For a founder it's usually neither. It's a calendar problem you inherited and never renegotiated.
Reconciling money that arrives through a payment processor
Let's be honest, most guides on month end close reconciliation assume money arrives by invoice and lands in the bank at face value. For a founder-run business it usually doesn't. It arrives through Stripe, PayPal or Razorpay, in a lump, after fees. Where does that go wrong?
Why the payout figure is not your revenue
A processor payout is net: gross customer charges minus platform fees, minus refunds, minus any chargebacks and reserves. Book the payout as revenue and you understate both income and costs, so your margin looks better than it is. It's the most common revenue error in founder-kept books, and it survives for months because the bank reconciles perfectly.
The four-part split: gross charges, platform fees, refunds, chargebacks
Here's what that actually looks like. Every payout decomposes into four things, and your ledger needs all four.
Gross charges are revenue. Platform fees are an operating expense (they are not a revenue deduction). Refunds reduce revenue in the period they're processed, and chargebacks reduce revenue while often carrying a separate dispute fee.
Your processor publishes a monthly summary carrying all four figures, and that report is the external reference your reconciliation ties back to. Stripe calls its version the balance summary report.
Payouts that straddle the month end
Think of it this way. Charges taken on the 30th and paid out on the 3rd belong to the earlier month; the cash belongs to the later one. That timing difference needs somewhere to live, which is what a clearing account is for. Skip one and you end up either restating revenue or quietly deciding revenue equals cash.
The clearing-account method, in four moves
Our recommendation is to create a processor clearing account on the balance sheet. Post gross charges as revenue with the contra side to clearing, then post fees, refunds and chargebacks against clearing as the report shows them. When the payout hits the bank, code it as a transfer from clearing. The closing clearing balance should equal money the processor is holding but hasn't yet paid you.
Locking the period: the step that makes it a close
Period lock accounting sounds like a housekeeping detail. It isn't. Locking is the control that converts a set of numbers into a closed month, because it fixes what those numbers were on the day you signed off. Why does that matter so much for a founder?
An unlocked period is a draft, not a close
Think of it this way. If March can still be edited in July, March's profit figure is a current opinion rather than a historical fact. Every downstream number inherits that instability: comparatives shift, the tax computation moves, and any variance analysis you did becomes unreproducible.
Setting a closing date in QuickBooks Online
In QuickBooks Online the control is called a closing date. You set the date through the account and settings area, and you can attach a password so any attempt to post or edit on or before that date requires it. QuickBooks Online Advanced carries the same mechanism (so a monthly close is supported there too).
Setting lock dates in Xero, and who is allowed to move them
In Xero the control is called a lock date, and it blocks postings dated on or before the date you set. Setting and moving lock dates is gated by role: you need the Adviser user role. Can you still post into the current period once a lock date is set? Yes, because the lock applies on or before the date, not to the open month.
That role gate stops being a software setting once somebody outside the company holds the books. It belongs alongside the clauses that decide who can change what.
Locking a period in Zoho Books
In Zoho Books the control is called transaction locking. You set a lock date, and transactions recorded before that date can no longer be added, amended or deleted until the lock is lifted. As with the other two, the lock only means something if the permission to lift it is restricted to one person.
What to do when your software only supports an annual close
Some desktop packages were built around an annual closing entry and never offered a monthly equivalent. The workaround is procedural: run the checklist monthly, export a PDF trial balance and statements on the day you finish, file them in the month's evidence folder, and restrict permissions on prior periods. But it's weaker than a software lock, and considerably better than nothing.
Digital-filing mandates quietly abolished the retrospective close
Here's the part almost nobody plans for. India's audit-trail requirement under the proviso to Rule 3(1) of the Companies (Accounts) Rules, 2014 means accounting software has to record an audit trail of every transaction and keep an edit log that cannot be disabled. A bookkeeper who reconstructs three months in one sitting produces a record whose edit history shows exactly that.
The second-order effect is uncomfortable. Your process is now visible to anyone who reads the log, not just your numbers. And the only way to control what that log says about you is to close and lock on time, every month.
The acceptance test: six conditions a closed month has to pass
Everyone tells founders to review and approve at the end of the close, and almost nobody says what passing looks like. So here it is as a test you can run rather than a feeling you have to trust. A month is closed when all six conditions below are true, and it is not closed when any one of them isn't.
The six conditions, and what to do when one fails
Condition | What passing looks like | What to do if it fails |
1. Every bank and card account is reconciled to its statement | Closing balance in the ledger equals the closing balance on the issued statement, for every account | Find the difference before doing anything else. Do not adjust the ledger to match; identify the transaction |
2. Suspense is cleared to zero | The suspense account balance reads zero on the last day of the month | Code each remaining item to a defensible account and write one line explaining the reasoning |
3. Control accounts agree to an external reference | Sales tax or VAT, payroll liabilities and loan balances each agree to a return, a payroll report or a lender statement | Reconcile the control account before filing anything that depends on it |
4. Payables and receivables sub-ledgers agree to the general ledger | Both sub-ledger closing balances equal their control-account balances, and the ageing reports look sane | Check for transactions posted straight to the control account instead of through the sub-ledger |
5. The period is locked | A closing date or lock date is set and the permission to move it is restricted | Lock it now. If you can't, export the statements to PDF and file them as the frozen record |
6. Every material variance is explained, not just noticed | Each line that moved beyond your threshold has one written sentence saying why | Investigate until you can write the sentence. "Probably timing" is not an explanation |
How to run the whole test in under thirty minutes
The short answer is that it fits into half an hour. Open the trial balance and the prior-month comparison side by side, then work the six conditions in order. Conditions 1 to 4 are arithmetic and take about five minutes each once your reports are saved as favourites. Condition 5 is a settings check, and condition 6 is the only one that takes thinking.
Why bother, if you already ran the checklist? Because the checklist proves you did the work and the test proves the work came out right. Those are different questions, and only the second one protects you.
Running the same six conditions against someone else's first close
Bottom line: the test is portable. If a bookkeeper, an accountant or an outsourced finance team hands you their first closed month, run exactly these six conditions against it before you accept it. Ask for the reconciliation reports, the suspense balance, the control-account agreements and the lock date, and ask for one sentence per material variance.
Our recommendation is to make that explicit rather than discovering it in month three. Acceptance conditions like these belong written into the scope so nothing slips, alongside the delivery dates.
Closing alone: controls when you are the entire finance function
Every serious month-end guide in circulation distributes the close across a team: a bookkeeper posts, a senior reviews, a manager approves, a partner signs off. Some publish a four-role responsibility matrix and state plainly that the close breaks without a review layer. All of that is true, and none of it is usable when the entire finance function is one person with a laptop. So what do you do instead?
The control you structurally cannot have
Here's the uncomfortable part. Independent review is the last control in the close, and it's the only one you cannot perform on yourself. That isn't a discipline failure. It's arithmetic: the person who made the entry is the person checking the entry, and the mind that missed something at 11pm on Tuesday will miss the same thing at 11am on Wednesday for exactly the same reason.
The ride-hailing company that published a margin figure with an extra zero had a finance department, external auditors and a disclosure process, and the review layer still didn't catch it before the market did. If that can happen with all that machinery, the question for a solo founder isn't whether to worry. It's which compensating controls actually work.
Time-separated self-review: close Monday, review Wednesday
And the cheapest substitute for a second person is a second occasion. Finish the close on one day, leave it alone completely, then review it cold two days later with the statements open and the checklist shut. You're not re-running the work; you're reading the output as a stranger would, which is a genuinely different cognitive task.
Founders who ask for a step-by-step guide so they can do their own bookkeeping usually get the steps and never get this. The gap between doing and reviewing is where the value sits, and separating them in time is the only lever a solo operator has.
A fixed variance threshold removes the judgement from the judgement call
A smarter strategy is to decide once, in writing, what counts as material: a percentage, an absolute figure, or whichever is larger. Then any line that moves past it gets investigated, whether or not you feel like it needs it. But the point isn't the number you pick; it's that you pick it before you see the results, so the threshold can't quietly adjust itself to how tired you are.
Is being fast at the close always trading against accuracy? Not when the threshold is fixed in advance. A fast close with a written variance rule beats a slow close where every judgement gets made fresh.
Period locking as a commitment device, not a formality
Now, here's where it gets interesting. For a founder, the lock date does a job it doesn't do inside a finance team. In a team, it stops other people editing history.
Alone, it stops you, which converts "I'll tidy that up later" into a decision you have to consciously reverse. That friction is the control.
The three checks worth buying an outside pair of eyes for
And if you buy one hour of external review a quarter, spend it on three things: the revenue recognition treatment, the payroll and contractor liability balances, and anything sitting in a balance-sheet account that hasn't moved in three months. Which kind of professional you're buying (a bookkeeper, an accountant, a controller or a fractional finance lead) is a separate question from how the close is run, and it belongs to the decision about who should own your books rather than to the procedure itself.
Automation moved your job from doing the close to reviewing it
Here's the shift nobody told founders about. When bank feeds auto-match and coding rules fire automatically, the failure mode inverts: errors stop looking like gaps and start looking like completed work. "Did it reconcile?" is now a question your software will almost always answer yes to, because it matched something to something.
So the valuable monthly skill has quietly changed. It's no longer categorising transactions, which the software does. It's knowing which three numbers to interrogate, and being willing to open the ones that look fine.
The same close, three markets
The accounting steps in a monthly close are universal. The compliance overlay on top of them is not, and the overlay is what actually sets your calendar. A founder with a Delaware C-corp, a UK VAT registration and an Indian delivery entity is currently served by three separate articles that don't know the others exist. So what changes, and what doesn't?
What is universal, and what is not
Record, reconcile, adjust, agree, report, lock: those six verbs are the same in every market on earth. What differs is retention (how long you must keep what the close produces), indirect tax (what you file and when), withholding (what you deduct and deposit), the filing-software mandate, and whether your software has to keep an edit log. Three of those five carry a hard date, which is why the overlay drives the calendar rather than the other way round.
One close, three compliance overlays
Universal close step | United States | United Kingdom | India |
Record retention | Books of account plus the software-maintained audit trail | ||
Indirect tax | State sales tax, nexus-dependent, with no single national date | VAT return plus digital record-keeping under Making Tax Digital | |
Withholding | Payroll tax deposits and the contractor information-return trail | Payroll deductions and the payroll control-account check | |
Software and filing mandate | No general mandate to file company accounts through commercial software | ||
Audit trail / edit log | No general statutory edit-log requirement for private companies | Digital records required, no general edit-log mandate | |
What sets the date | Bank statements, payroll finalisation and processor settlement | The same three, plus the VAT return cycle | The same three, plus the 7th and the availability of the credit statement on the 14th |
United States: retention periods and the four-year employment-tax rule
For a US entity, how long you keep the records a close produces isn't a single number. Retention is keyed to the period of limitations that applies to the return those records support, and employment-tax records carry their own four-year rule: the IRS asks you to keep them for at least four years after the tax becomes due or is paid, whichever is later. Practically, the reconciliation evidence you file each month has a defined shelf life, and that shelf life runs longer than most founders assume.
United Kingdom: VAT, the control-account check, and digital record-keeping
A UK founder's close carries one extra mandatory step: agreeing the VAT control account before the return goes in. HMRC's VAT guide sets out the underlying rules that control account has to reflect, and Making Tax Digital already requires VAT-registered businesses to keep digital records and submit through compatible software rather than a spreadsheet and a web form.
India: the 7th, the 14th, and the audit trail
India's close is the most date-gated of the three, and two dates do most of the work. Tax deducted at source is payable by the 7th of the following month, section by section, which drags payroll and contractor coding forward into the first week. Then input tax credit can only be claimed where the invoice appears in the recipient's auto-generated statement, a condition set by section 16(2)(aa) of the CGST Act, and that statement becomes available on the 14th.
Add the audit-trail requirement obliging accounting software to keep an edit log, and the Indian close stops being a question of internal readiness. It's gated by an external portal and evidenced by your own software.
What changes for UK founders from April 2026
Two changes are already scheduled, and both reward planning now. Making Tax Digital for Income Tax starts on 6 April 2026 for those with qualifying gross income above 50,000 pounds, stepping to above 30,000 pounds from April 2027 and above 20,000 pounds from April 2028, with digital records and quarterly submissions through compatible software. But the trap sits in the definition: qualifying income is gross turnover, not profit, so a founder running thin margins can cross the threshold without feeling remotely like a 50,000-pound business.
The second is filing. Under the Economic Crime and Corporate Transparency Act 2023, all companies including micro-entities and dormant ones will file annual accounts in iXBRL format through commercial software, and Companies House has confirmed the change for April 2028. And several secondary pages still publish the superseded 2027 date, which is worth knowing before you plan around it. Early signals point one way: a founder who has been closing "whenever" acquires four statutory checkpoints a year plus a software-mediated filing route, and both of those reward a locked monthly close.
What the close produces, and who reads it later
A closed month isn't a private administrative event. It produces documents other people will read, sometimes years later, usually at a moment when you'd rather they didn't have questions. Who are those people?
The four outputs of every close
Every close produces four things: a trial balance that balances, a profit and loss statement, a balance sheet as at the last day, and a cash flow statement. Read them in that order, because the trial balance tells you whether the other three are built on anything solid. Interpreting the cash flow statement is its own subject, and what the cash flow statement is actually telling you needs more room than a checklist can give it.
The four future readers of your books
Your accountant reads them at tax time, looking for the schedules and the evidence behind the adjustments. A lender reads them for consistency between periods (month to month, not just year to year). An acquirer reads them for whether the trail is locked and reconciled, because a locked, reconciled trail is what diligence reads. A tax authority reads them for whether the records exist at all, which is what guidance such as IRS Publication 583 sets out for US businesses.
Worth flagging what happens when that control layer is judged to have failed. A US-listed server manufacturer spent roughly six months out of filing compliance from August 2024, and its replacement auditor issued an adverse opinion on internal control over financial reporting. The monthly close is the smallest working unit of that control.
A one-page close policy you will actually follow
Frankly, this gets overlooked. Write down five things and nothing more: the day you start, the day you finish, your variance threshold, who (if anyone) reviews, and where the evidence lives. That's your month-end close policy. A one-page policy you follow beats a twelve-page one you never opened again.
How often to review the process itself
The practical reality is that twice a year is enough for most founder-run businesses, plus any time something structural changes: a new bank, a new processor, a new market, a first employee. And look at where the close actually stalled, not where you assumed it would.
Where the close is heading
The monthly close has changed more in the last fifteen years than in the fifty before it. So what's actually changing?
From statement-matching to review-and-approve
Between roughly 2010 and 2016, cloud accounting plus automated bank feeds moved reconciliation from statement-matching to review-and-approve, the biggest reduction in founder close effort in twenty years. But it created the modern failure mode: feeds that look reconciled because transactions were auto-matched and never reviewed.
Automation is eating the matching, not the judgement
Early signals suggest the pattern extends into anomaly detection and payables automation: the mechanical parts keep getting cheaper and faster.
Now, here's where it gets interesting. What has not moved is the judgement layer. Software can tell you two numbers agree. But it cannot tell you they agree for the wrong reason.
What gets scarce is review, not data entry
The cheap end of bookkeeping keeps getting cheaper, while the US Bureau of Labor Statistics projects employment in bookkeeping, accounting and auditing clerk roles to decline over the 2024 to 2034 decade, and the supply of new accounting graduates has kept shrinking, per the AICPA's 2025 Trends report. A founder can buy an inexpensive close but struggles to buy someone to check it. Which makes the six-condition acceptance test the highest-leverage thirty minutes in your month.
The tools that shorten a close
Four things reliably cut close time at a category level: automated bank and card feeds, coding rules for recurring transactions, a receipt-capture tool that files evidence as it arrives, and a reconciliation report saved as a favourite (so it opens in one click). Based on what we've seen, which products you choose matters far less than whether you set them up before the month ends.
When the close outgrows you
Three months in, some founders find the procedure fits and some don't. The honest signal isn't a transaction count or a revenue figure. It's that the close keeps landing on days you've already committed elsewhere, month after month.
Can a small business hand the monthly close to someone else? Yes, and the six-condition acceptance test is what you hand over with it, so a provider's first closed month is testable rather than merely trusted.
What this guide deliberately doesn't answer is what you should pay for bookkeeping, or when a founder should bring in outside finance support. Both turn on economics rather than procedure, and both belong to the question of whether the books should be handed to someone else at all.
If you'd rather the close simply ran. Outsource360's virtual bookkeeping and accounting service runs the procedure on this page for founder-led companies: reconciliations agreed to statement, sub-ledgers and control accounts tied back to the ledger, the period locked, and the four statements delivered on a fixed date every month.
Frequently asked questions
1. How long should a monthly close take for a small business?
Guides quote five to seven working days, and that's achievable for a simple business. The real constraint is dependency timing. Practitioners describing their own closes report anything from about two days to roughly 45 to 60 days at some privately-held companies.
2. What's the difference between the monthly close and the year-end close?
The monthly close is the base procedure: record, reconcile, adjust, agree, report, lock. The year-end close runs the same steps, then adds fixed-asset and depreciation schedules, stock counts, accrual true-ups, statutory accounts and the tax computation. One is read internally, the other externally.
3. What's the difference between bookkeeping and the monthly close?
Bookkeeping is the daily habit: coding transactions, raising invoices, filing receipts. The monthly close is the periodic control that checks the daily habit was right, then freezes the result. You can do bookkeeping every day and never close a month.
4. What is month-end reconciliation?
Reconciliation means agreeing a balance in your ledger to an external reference, and nothing else counts. For a bank account that reference is the statement the bank issued, closing balance included, not the feed your software imported. For a processor, it's the monthly fee report.
5. What are accruals and prepayments, and do I need them every month?
An accrual books a cost into the month it was incurred even though the invoice hasn't arrived. A prepayment spreads something already paid across the months it covers, such as an annual licence. We'd recommend posting them monthly if the amounts are material.
6. What reports should I have at the end of every close?
Four: a trial balance that balances, a profit and loss statement for the month, a balance sheet as at the last day, and a cash flow statement. Read them in that order, because the trial balance tells you whether the other three stand on anything solid.
7. What does a "closed" month actually mean?
A month is closed when six conditions hold: every bank and card account reconciled to its statement, suspense at zero, control accounts agreed to an external reference, sub-ledgers agreed to the general ledger, the period locked, and every material variance explained in writing.
8. How do I lock a period in QuickBooks Online?
QuickBooks Online uses a closing date, which you set in account and settings. You can attach a password so that posting or editing on or before that date requires it. Without the password the closing date acts as a warning rather than a barrier.
9. How do I set lock dates in Xero, and who can change them?
Xero uses lock dates, which block postings dated on or before the date you set. Setting and moving them is gated by role: the permission sits with users holding the Adviser role. Your close date is only as firm as your user-access list.
10. Can I close the books monthly in QuickBooks Desktop?
The honest answer is that desktop versions were built around an annual closing entry rather than a monthly one. The workaround is procedural: run the checklist monthly, export a PDF trial balance and the statements on the day you finish, and restrict permissions on prior periods.
11. How do I reconcile Stripe or another payment processor at month-end?
Four moves. Open a processor clearing account on the balance sheet, post gross charges as revenue against clearing, post fees, refunds and chargebacks against clearing from the processor's monthly report, then code the bank payout as a transfer from clearing.
12. What is a suspense account, and why must it clear to zero?
Suspense is temporary storage for transactions you couldn't identify when they landed. Used as a queue it's healthy. Left with a balance at month end it hides real items inside a placeholder, and a year-end balance is one of the first things an accountant queries.
13. What are control accounts, and how often should I check them?
A control account holds the summary balance for a group of transactions: sales tax or VAT, payroll liabilities, loans and any owner's or director's account. Check every one monthly, and agree each to an external reference rather than to your own workings.
14. Do I need to count inventory every month?
Only if you hold stock and its value is material to your numbers. A services business skips this step entirely. A product business almost never should, because unrecorded shrinkage, damage and miscounts flow straight into cost of sales and distort every margin figure.
15. Do I need to reconcile petty cash if I barely use cash?
If the amounts are negligible, a quarterly count with a written policy is a reasonable position. What isn't reasonable is claiming a monthly count you don't perform. Decide the cadence, write it into your close policy and hold to it.
16. How long do I need to keep the records a close produces?
In the US, retention is keyed to the period of limitations for the relevant return, with employment-tax records kept at least four years after the tax becomes due or is paid, whichever is later. In the UK, VAT records and accounts are retained under HMRC's VAT guide. In India, books of account sit alongside the software-maintained audit trail.
17. What are the most common month-end closing mistakes?
Five recur at every size of business: adjusting before reconciling, reconciling to the imported feed instead of the issued statement, booking a processor payout as revenue, leaving a balance in suspense, and never locking the period. And the fifth turns the other four into arguments you are still having a year later.
18. Can a founder run the close alone, or do I need a second reviewer?
You can run every step alone. What you cannot do alone is independently review your own work, which is a structural limit rather than a discipline problem. The practical reality is that you compensate with time-separated self-review, a written variance threshold fixed in advance, and a lock date you treat as final.
Disclaimer
This article is for educational and general business information purposes only and does not constitute professional accounting, tax or legal advice. Accounting requirements, filing dates, retention periods and record-keeping obligations vary by jurisdiction and by entity type, and they change over time. For guidance specific to your situation, consult a qualified professional before acting.
Comments