Someone good has become available, or the team is stretched, and you want to know whether you can carry another salary. The honest answer most owners give themselves is a feeling about the bank balance.

The bank balance is the wrong instrument. It includes money you owe to suppliers, VAT you’re holding, payroll going out next week, and it excludes anything a customer hasn’t paid you yet. A business can look comfortable on Tuesday and be short by the end of the month without anything going wrong.

What you need instead is dull and specific: last three months of gross margin, a debtor list you believe, creditors with dates, and payroll as a percentage of revenue. That’s four figures. If getting them takes a fortnight of catching up, the bookkeeping has been outgrown.

In practice the hire either gets made on optimism or gets postponed on nerves. Both cost something. One of them costs more than the other, and you won’t know which until later.

Year-end accounts filed nine months after the period they cover satisfy Companies House and HMRC. They tell you nothing you can act on. By then the pricing was wrong for nine months, or the job you thought made money didn’t.

The sign here is timing rather than accuracy. Ask yourself a plain question: on the fifteenth of this month, could you see last month’s results? If the answer is no, every decision you make in between runs on memory.

Owners usually compensate by building their own spreadsheet. That’s a reasonable response and a bad long-term answer, because the spreadsheet is a second set of records with no reconciliation behind it. Two versions of the truth is worse than one late version.

We think a management pack landing on an agreed date each month is the single change that moves a business from reacting to deciding. More detail on that in why your numbers arrive too late to help.

This is the most common one we meet. The profit and loss says the year is going well and the current account disagrees. The owner concludes the accountant is wrong, or that profit is a theory.

Usually the money is sitting in three places. Unpaid customer invoices that nobody is chasing to a routine. Stock or work in progress bought and not yet billed. And tax set aside in principle but not in practice.

The awkward version of this is work you’ve delivered and never invoiced at all. Reconciling a sales ledger against delivery records for one owner-managed trading business turned up £20,000 of goods supplied and never billed. Nobody had stolen anything, and nobody was careless. The invoicing simply sat outside any routine, so it happened when someone remembered.

Once invoices were raised and collected, invoicing moved into month-end and stopped depending on memory. That’s the pattern: not a clever fix, a boring routine applied consistently.

A bank wants management accounts for a lending decision. A landlord wants covenant figures. A prospective buyer or an investor asks for a breakdown by service line. A funder wants to see where their money went.

You can usually get there eventually. The question is what eventually costs. Deals slow down, lending conversations stall, and the person on the other side draws their own conclusion about how the business is run.

There’s a second version of this inside the business. Your accountant asks why a particular cost jumped, and nobody knows. Asking that sort of question is how we found a not-for-profit had over £200,000 of funding with the paperwork never completed. The money had been awarded. Nothing had been claimed.

If you can’t answer an outsider’s question in the meeting, you’ve outgrown records that exist only to support a tax return.

The last sign is structural. Ask who reconciles the bank, who chases debtors, who checks the supplier statements, who runs payroll and who reviews it. If the answer to four of those is the same person, and that person also runs the business, the finance function is a habit rather than a process.

Habits break when you’re busy, which is exactly when the numbers matter most. They also break on holiday, and they don’t survive growth.

What replaces it isn’t complicated. A written order of what happens when, an owner for each task, and a date the reporting lands. We spend the first ninety days on exactly that: complete and reconciled records first, then routines and controls, then monthly management information you can act on. Software helps with the mechanics, but a good system on top of incomplete records produces tidy reports about nothing.

If two or more of these five signs describe your month, the bookkeeping has been outgrown. The interesting question is what you’d do differently next quarter if the figures were in front of you.