From cash drawer to clean books: a month-end that takes an hour

At most sports facilities in Bangladesh, month-end is an archaeology project. Someone collects receipts from a drawer, reconciles a bKash statement against a notebook, guesses at which expenses belong to which ground, and hands an accountant a pile of paper. The accountant produces a number three weeks later. Nobody entirely believes it, including the accountant.

This is not because facility owners are disorganised. It is because the recording happens after the fact, and anything recorded after the fact is a reconstruction.

The four things that make month-end slow

Money is recorded once, not twice

A single-entry list — "7 PM booking, ৳2,500, cash" — tells you what came in and nothing about where it went. Double-entry records both sides of every transaction: cash went up, revenue went up. It sounds like extra work, and when done by hand it is. When the software does it as the sale happens, it costs nothing and the accounts balance by construction.

Cash and bank are the same line

When cash at the desk, bKash, card and bank transfer all land in one "income" figure, you cannot reconcile any of them. Each payment method needs its own account, so that at day close you can compare what the drawer holds with what the system says it should hold — and see the variance while the day is still fresh enough to explain it.

Expenses arrive without a home

A ৳7,400 electricity bill for a facility with three grounds is not one number, it is three. If it is not split when it is entered, someone splits it from memory later, and the per-ground profit you thought you had was never real.

Nothing is ever final

If March can still change in June, then March is not a report, it is a draft. Without a genuine close, every month stays provisional and nobody can safely act on last quarter’s numbers.

What has to be true instead

• Every transaction posts itself, in the moment, from the same action that took the money — no re-keying, no batch entry at month end.
• Each payment method has its own account, so day close is a five-minute count rather than a monthly investigation.
• Expenses are tagged to a facility when they are entered, by the person who knows the answer.
• Payroll posts like everything else: gross, deduction, net, split across the venues a person actually works at.
• The month can be closed, and a closed month stops accepting entries.

What "closing the month" should feel like

Open the profit-and-loss for the period. Look at whether anything is obviously wrong — a category with a number you do not recognise, a facility with no expenses at all. Reconcile the bank statement against the entries, which mostly auto-match, leaving you a handful of exceptions to work. Then close the period.

That is the hour. The reason it is an hour is that the bookkeeping already happened, continuously, as a by-product of running the business.

The late invoice, which is the real test

A supplier’s bill arrives in the second week of the following month, dated to the month you just closed. This single scenario separates real accounting from a report screen.

The wrong answer is to reopen the closed month and slip the entry in, because now the P&L you already showed your partners is different from the one in the system, and no one can tell which version anyone was looking at. The right answer is the one auditors expect: post it into the current open period as an adjustment that references the original date. The closed month stays exactly as reported, and the correction is visible as a correction.

A closed month that can still change was never closed. It was just quiet.

This is why Sportify locks a closed period at the database level and records late items as adjustments in the next one. It is also why the ledger only ever appends: a mistake is fixed with a reversing entry rather than an edit, so the history of what you reported survives intact.

Where to start if you are doing this by hand

Even before you change software, two habits pay for themselves immediately: separate your payment methods into their own lines, and count the drawer against an expected figure at the end of every day. The first makes reconciliation possible at all. The second turns a monthly mystery into a daily two-minute question, asked while people still remember the answer.