What a clean monthly close actually looks like

Most founders don't get bad numbers because their bookkeeper is careless. They get bad numbers because the month was never actually closed. The bank feed imported, the categories looked roughly right, and someone hit a button. That is not a close. That is a snapshot of an unfinished room.

A monthly close is a short, repeatable set of checks that confirm one thing: every dollar that moved is recorded, in the right month, in the right place, and nothing is hiding. When that's done, you can read your profit and loss statement and trust it. When it's not, you end up explaining away the numbers in every board update, which is exhausting and slowly erodes your own confidence in the business.

Here is what a clean close actually involves, in the order we do it, and the cadence we aim for: books locked by the fifth business day of the following month.

Reconcile every account that holds money

Reconciliation means proving your books match the real world. You take the bank's ending balance for the month and confirm your accounting software arrives at the exact same number, transaction by transaction. Not close. Exact. A $40 difference is not rounding; it's a missing or duplicated entry, and it will compound.

Do this for every account that touches cash: each checking and savings account, every credit card, and any payment processor like Stripe or PayPal that holds a balance before paying out. Processors are where teams get burned, because a $10,000 batch of sales might land in the bank as $9,710 after fees, and if you only record the $9,710 your revenue is quietly understated by the fees every single month.

The test for a real reconciliation is simple. Pull the statement, match it to the books, and confirm the difference is zero before you move on.

• Checking and savings: ending balance ties to the penny
• Each credit card: statement balance matches the books
• Stripe / PayPal / Square: gross sales recorded, fees booked separately
• Loans and lines of credit: principal and interest split correctly
• Any difference investigated, not plugged with a balancing entry

Get the cutoff right

Cutoff is just making sure each transaction lands in the month it belongs to, not the month the cash happened to move. This is where the accrual idea earns its keep. Accrual accounting records revenue when you earn it and costs when you incur them, regardless of when money changes hands.

A concrete example. You send a client a $6,000 invoice on May 28 for work delivered in May, and they pay on June 12. On a cash view, May shows nothing and June shows $6,000, which makes May look weak and June look great for no real reason. With proper cutoff, the $6,000 sits in May where the work was done, and June correctly shows the cash arriving against a balance you already booked. Get this wrong across a dozen invoices and your monthly trend line becomes fiction.

Run the same logic in reverse for bills. A $1,200 contractor invoice for May work that arrives in your inbox on June 8 still belongs to May.

Book the accruals you can't see in the bank feed

Some real costs never show up as a clean transaction in the month they apply to, so you have to add them on purpose. These are accruals and prepaids, and skipping them is the most common reason a profit number looks better than the business actually performed.

Three that matter for almost every small company: payroll that straddles the month-end, so the days worked in May get expensed in May even if payday is June 3; an annual software bill paid once but spread across the twelve months it covers, so a $12,000 yearly subscription hits the books at $1,000 a month rather than a single ugly spike; and any recurring bill you know is coming but hasn't arrived yet, like the cloud hosting invoice that always lands a week late. None of these are optional if you want a profit number you can defend.

The owner review, and landing it by day 5

The final step is a human reading the numbers before they're called final. Not re-doing the bookkeeping, just sanity-checking it. The fastest version: open this month's profit and loss next to last month's and look for anything that moved more than, say, 20 percent. Rent jumped from $4,000 to $9,000? Either you actually moved offices or a vendor got coded to the wrong account. Revenue flat but you closed three new clients? Something didn't get invoiced. The review catches the errors that reconcile perfectly but still don't make sense.

On cadence, aim to lock the books by the fifth business day of the following month. That's realistic for most small teams, and the structure that gets you there is boring and effective: days 1 to 2 for bank and card reconciliations, day 3 for cutoff and accruals, day 4 for the owner review and fixing whatever it surfaced, day 5 to finalize and lock the period so no one quietly edits a closed month later. Lock matters as much as speed. A month that keeps changing was never really closed.

If your close regularly slips past the tenth, it's usually one of two things: a bank or processor that never quite reconciles, or accruals nobody owns. Fix those two and the whole month tightens up.