Most month-end close checklists tell you to "automate everything." That's the marketing version. The honest version is that some close steps automate cleanly, some need a human reviewing exceptions, and some are pure judgment calls software can't make. If you treat all 12 steps as equal, you'll either over-promise automation to your firm or under-use the tools you've already paid for.
This checklist tags every step as Manual, Automatable, or Already Automated. It's part of the broader bookkeeping automation framework Growthy uses to evaluate what's worth automating across a multi-client portfolio. Each step shows the time it took manually, the time it takes with automation, and where the judgment line sits. Use it to plan your close week, train staff, and decide which steps are worth investing in tooling for.
What can you actually automate at month-end close?
About 5 of the 12 standard close steps automate cleanly: bank and credit card reconciliations, recurring depreciation entries, period locking, and AI categorization with bookkeeper review. Another 5 stay manual because they require judgment a model doesn't have: aging reviews, inventory counts, fixed-asset capitalization, accruals, and intercompany transfers. Tax-basis adjustments and AP follow-ups land in between. A bookkeeper running 15 clients can cut close time from about 75 hours to about 26 hours per month using the automatable steps, without giving up control over the judgment calls.
Three labels appear in the checklist below. Each means something specific.
Already Automated means the software does the step end-to-end without human review. Recurring depreciation journal entries fall here. So does the period lock. The bookkeeper presses one button or sets one schedule and the work happens.
Automatable means the software does the work, but a human reviews flagged items. Bank reconciliation matches 90% of transactions automatically; the bookkeeper reviews the 10% that didn't match cleanly. AI categorization codes most lines; the bookkeeper reviews the ones with low confidence scores. Same pattern: software does the volume, human owns the exceptions.
Manual means there's no automation that can do the step well. Accruals require a judgment about what was earned but not invoiced. Intercompany transfers need entity context the system doesn't have. Inventory adjustments need someone to count what's on the shelf. These steps don't get easier with better software; they get easier with better process.
The split matters because vendor pitches blur the three. If a tool sells "automated close," ask which of the 12 steps it handles, and which label it earns on each.
Here are the 12 steps in the order most firms run them, with the label and time estimate for each. The time estimates assume one client of typical small-business complexity (under $5M revenue, single entity, 200-500 monthly transactions).
That's the full close. Five Automatable, two Already Automated, five Manual. The split is the same whether you're on QBO with bank rules, Xero with manual categorization, or Growthy with AI categorization. What changes is how much time the Automatable steps take.
Here's how the 12 steps map to a typical 4-day close week. The day labels assume a Monday close start with a Thursday wrap.
Day 1 (Monday): reconciliations and categorization. This is the automation-heavy day. Bank and credit card reconciliations run, and the AI categorization queue fills with flagged items. The bookkeeper reviews exceptions and clears them. If the close is going to land on time, this day moves fast.
Day 2 (Tuesday): aging and inventory. AR aging, AP aging, and inventory adjustments. All three are judgment days. The system tees up the data; the bookkeeper makes decisions. Plan for client emails on collectibility and vendor questions.
Day 3 (Wednesday): fixed assets, depreciation, accruals, intercompany. The middle of the close. Capitalize what needs capitalizing, post the recurring depreciation, book accruals, and clear intercompany. Multi-entity clients eat most of this day.
Day 4 (Thursday): tax-basis and close-lock. Apply any tax-basis adjustments needed for the period (often skipped if year-end only), then close and lock. Send the close package to the client or partner. Done.
Solo bookkeepers running 5-15 clients usually batch this into one or two days a month rather than spreading across the week. The order still matters: reconciliations and categorization first, judgment calls second, lock last.
For the steps tagged Automatable or Already Automated, here's what Growthy does differently from a stock QBO + bank rules setup.
Auto-bank-rec runs nightly, not on demand. When the bookkeeper signs in Monday morning, the rec is done and exceptions are queued. They start at the exception list, not the matching screen. That's the 30-minute-vs-90-minute difference on step 1.
Confidence scoring routes the queue. AI categorization assigns a confidence score to every transaction. Above 90% goes through automatically. Between 70-90% lands in a "review recommended" tier. Below 70% lands in "must review." The bookkeeper sees the must-review queue first and works down. No more clicking through 487 transactions one by one to find the 15 that needed attention.
Recurring journal entries post on schedule. Depreciation, prepaid amortization, and any other recurring entries post automatically when the schedule fires. The bookkeeper gets a notification with the entry posted and a one-click "review and approve" if firm policy requires.
Period lock is one click with audit trail. After step 12, the bookkeeper clicks lock. The system captures who, when, and which prior-period edits get blocked. If a partner needs to reopen the period for a correction, that's logged too.
None of this magic. It's just removing the busywork from the steps that don't need a human, so the human can spend time on the steps that do.
The math at scale is what makes the close-automation investment worth tracking. Here's a typical 15-client solo bookkeeper, comparing manual close to automated close.
Manual close: 15 clients × 5 hours per close = 75 hours per month. That's almost two full work weeks gone to close work, leaving little time for advisory, sales, or rest. Most solo bookkeepers hit a wall here around 12-13 clients because the math stops working.
Automated close: 15 clients × 1.75 hours per close = about 26 hours per month. Same client load, same quality, about a third of the time. The savings come almost entirely from steps 1, 2, and 3 (reconciliations and categorization), which add up to about 3 hours per client manually and about 1 hour per client automated.
Capacity recapture: 75 - 26 = 49 hours per month. At a $75/hour bookkeeper rate, that's about $3,675/month or $44,100/year of recovered capacity. The bookkeeper either takes more clients (the solo bookkeeper stack scaling to 15 clients walks through how) or repurposes the time into higher-margin advisory work.
The investment to get there is a stack cost of around $200-350/month for a 15-client setup. Net: about $40,000 of recovered annual capacity for around $4,000 of tooling. CPA firms running similar engagements see comparable math; scaling into a CPA-firm bookkeeping engagement covers the practitioner version. For more on the underlying time math, the real cost of manual bookkeeping breaks it down by activity.
Can I customize the checklist for my firm's process? Yes. The 12 steps are the standard small-business close pattern, but most firms add or remove steps. Some add a sales-tax accrual review (step 9.5). Some skip inventory entirely. The labels (Manual, Automatable, Done) stay the same regardless of step order; what changes is which tools you use to handle the Automatable ones.
What about multi-entity consolidations? The 12 steps still apply per entity, but consolidation reporting is a separate workflow. Growthy handles multi-entity natively (separate books per entity in one workspace), and consolidation reporting is on the roadmap for late 2026. Today, multi-entity close means running the 12 steps per entity, then exporting trial balances to a consolidation tool (Fathom, Reach Reporting) for the consolidated package.
Does Growthy enforce close-period locking? Yes. After step 12, the period locks and back-dated edits are blocked. Any reopen action (typically a partner override for corrections) is captured in the audit trail with timestamp and user. This matters for compilation and review engagements where audit-trail integrity is part of the workpaper.
How does the AI handle one-off vendors that only appear during close? Year-end accruals, annual insurance prepaids, and other once-a-year items often have no prior history to learn from. The AI flags them as low-confidence (under 70%) and routes them to the bookkeeper review queue. After the bookkeeper codes them once, the system learns the pattern for next year. Same with new clients in their first close: confidence scores start lower and improve over the first 2-3 close cycles.
What if the bank feed is broken on close week? Same answer as without automation: you import a CSV or PDF statement and reconcile from that. The auto-rec works against any feed source, so a CSV import still gets matched. Manual fallback exists; you don't lose the close because Plaid had a bad day.
Growthy is bookkeeping software, not a CPA firm. This content is educational, not professional advice. Full disclaimer.
Related: Bookkeeping Automation for Multi-Client Bookkeepers, Automated Bank Reconciliation, Batch Deposit Matching, Bookkeeper Automation Stack
Free during alpha. Read-only access. You review every sync.
CPA firm partner who got tired of watching bookkeepers click categorize 500 times a day. Built Growthy to fix it.
View all articles →Growthy is dedicated to helping businesses of all sizes make informed decisions. We adhere to strict editorial guidelines to ensure that our content meets and maintains our high standards.
When QBO stops scaling for bookkeepers running 15+ client portfolios. 5 breakpoints past 15 clients, honest comparison vs Growthy, Xero, Puzzle, FreshBooks, plus a 30-day migration overview.
Bank rules vs AI categorization across 8 dimensions. Hybrid stack wins: rules for high-confidence repeat vendors (Stripe, payroll, recurring SaaS), AI for everything else. Real numbers.
