Now in open beta — close the books in 2 days, not 2 weeks.Read the case study →
Business · January 15, 2026 · 11 min read

Monthly close process: cut 18 days to 3

Most SMBs close their books 15 to 25 days after month-end. The fast ones close in 3. The difference is not horsepower, it is discipline plus the right tooling. Here is the 12-step cycle that gets you there without skipping work or compromising audit quality.

Why most closes take three weeks.

A typical SMB close lags because of three failure modes. First, cut-off discipline is missing: invoices keep arriving for the prior month a week into the new month, and AP keeps re-opening the period. Second, reconciliation is a marathon at the end instead of a habit during the month. Third, review and adjusting entries pile up because no one has time during the month to handle them as they appear.

The fix is not heroics. It is moving work earlier, parallelizing what does not depend on the period being closed, and using automation for the repetitive bits. Done well, a 4-person finance team can close a £8M revenue business in 3 working days every month.

The 12-step accelerated cycle.

These twelve steps run roughly in parallel across days 1 to 3 of the new month. Steps 1 to 4 happen on day 1. Steps 5 to 8 on day 2. Steps 9 to 12 on day 3 with the final P&L review. The key insight is that once cut-off is enforced, most of the work can run as soon as bank statements are available, not after every last invoice arrives.

A worked example. A 3-branch UK trading company with £7M annual revenue ran a 22-day close in 2025. After moving to the 12-step cycle on Nonari, they hit a 4-day close in month two and a 3-day close in month four. The compounding factor was the AI auto-coding: by month four the classifier was hitting 78 percent precision on routine entries, freeing the controller to focus on the 22 percent that needed judgement.

Day 1Lock + auto-codeDay 2Reconcile + adjustDay 3Review + sign-off
Three days, twelve parallel steps. Most work runs the moment bank statements land.
  • Day 1: lock prior period, pull bank statements, run auto-code
  • Day 1: AP cut-off review, AR aging, inventory cut-off
  • Day 2: bank reconciliations, intercompany, payroll JE
  • Day 2: accruals, prepayments, depreciation
  • Day 3: review queue, adjusting entries, branch P&L
  • Day 3: consolidated P&L, variance commentary, sign-off

Step 1 to 4: lock and load on day 1.

Step 1 is hard cut-off. The prior month is locked at 11:59 PM on the last calendar day. New entries dated to the prior month require an explicit unlock action recorded in the audit log. Step 2 is bank statement pull: every account, BAI2 / MT940 / OFX / CSV, into the reconciliation queue. Step 3 is auto-code on the imported transactions: Nonari processes these in seconds for repeating patterns. Step 4 is AR aging snapshot and AP cut-off review.

A common mistake is letting the prior period stay open "just in case" a late invoice arrives. The cleaner approach: post late invoices in the current period with a clear date narration, and only re-open the prior period for material adjusting entries that would distort comparisons. Material is usually defined as 1 percent of revenue or £20,000, whichever is lower.

Step 5 to 8: reconcile and adjust on day 2.

Step 5 is bank reconciliation. With AI matching, this collapses to exception review only. Step 6 is intercompany reconciliation if you have multiple legal entities. Step 7 is the payroll journal entry with PAYE / national insurance / pension reconciliation (or the local equivalent — US 941, Australian PAYG, Canadian CPP/EI). Step 8 is accruals, prepayments, and depreciation.

Depreciation should never be manual at this stage. Asset registers with monthly depreciation rules post automatically. Prepayments unwind on schedule. Accruals follow recurring templates. The only manual work on day 2 is exception items that did not match a rule or template.

Step 9 to 12: review and report on day 3.

Step 9 is the AI review queue: every transaction the classifier was less than 85 percent confident on, surfaced for human decision. Step 10 is manual adjusting entries: bad debt provisions, inventory write-downs, foreign exchange revaluation. Step 11 is the branch-level P&L pull: each branch reviewed for sanity. Step 12 is the consolidated P&L plus variance commentary versus prior month and budget.

The branch-level P&L is where many issues surface. A branch that suddenly has 8 percent gross margin compression has either an inventory cost issue, a discount-policy issue, or a theft issue. Catching that on day 3 of the new month, not 18 days later, is the difference between a fixable problem and a quarter-long bleed.

What changes on Nonari versus a manual workflow.

Three things compress dramatically. Auto-coding removes 60 to 75 percent of the typing work. Bank reconciliation moves from a 4-hour task to a 30-minute exception review. Branch P&L generation moves from a half-day Excel exercise to a single click. The remaining work — judgement, review, commentary — is exactly where you want a finance person spending their time.

Operators who have made this transition report a second-order benefit. With closes happening fast, monthly P&L commentary becomes a real management tool, not an autopsy. By the time a manager reads March numbers, May is already half over. Closing in 3 days means March numbers land on April 3, and you can act on April 4.

Common pitfalls and how to avoid them.

Pitfall one: letting AP keep flowing into the prior period. Fix: hard cut-off plus a "post to current with date narration" rule for late invoices. Pitfall two: trying to reconcile everything at once. Fix: bank rec runs daily during the month, so the closing rec is a verification, not a reconstruction. Pitfall three: no review queue discipline. Fix: queue items have an SLA — 24 hours for routine, 4 hours for material.

Pitfall four, and the silent killer: nobody owns the close. Assign a single named owner, even in a small team, with a daily standup during close days. The standup is 10 minutes and surfaces blockers fast. This is the highest-leverage change for most SMBs and it costs nothing.

Frequently asked

Common questions.

Is a 3-day close realistic for a single bookkeeper SMB?

Yes for businesses up to about £4M revenue with mostly routine transactions. The bookkeeper handles steps 1 to 8 across two days using automation, and the owner reviews the queue and signs off on day 3. Above £4M with multiple branches you typically need a 2-person finance team to hit 3 days reliably.

What if late invoices arrive after we have closed?

Post them in the current period with a clear date narration. Only re-open the prior period for material adjusting entries — typically 1 percent of revenue or £20,000, whichever is lower. Re-opening for small items destroys close discipline and adds no information value.

How long does it take to get from a 20-day close to a 3-day close?

Most teams hit a 5-day close in month two of using disciplined cut-off plus automation, and a 3-day close by month four once the AI classifier has learned their patterns. The bottleneck after that is usually review queue SLA, which is a process issue, not a tooling one.

Do we still need a year-end audit if monthly closes are tight?

Yes if you are above local audit thresholds (Companies House in the UK, statutory audit triggers elsewhere). A fast monthly close gives you decision-quality numbers but does not replace independent audit. The good news: tight monthly closes make the year-end audit faster and cheaper because fewer adjustments and reconciliations are pending.

Try nonari

Put your books on autopilot.

Free to start. No credit card. Bring your books, kick the tires, export everything if you decide to leave.