How much bookkeeping should a founder still be doing?

In the earliest months, often yes — volume is low and doing it builds a useful instinct for how money moves through the business. It stops being sensible once the work is recurring rather than occasional, because recurring execution attached to a founder becomes the constraint on everything else. The durable split is that a founder owns decisions: the cash position, releasing payments, the margin story, and any concession. The recording, matching, chasing and formatting should not stay.

There is no correct number, and any figure quoted as a benchmark is invented. The useful measure is composition rather than duration. Time spent deciding — approving what leaves, resolving a disputed invoice, choosing whether to grant terms — is well spent at almost any length. Time spent categorising, matching, chasing documents or rebuilding a spreadsheet is a cost regardless of how short it is, because it produces nothing that lasts beyond the task.

Four. Knowing the cash position net of committed money, so no decision is made against a number that is already spent. Approving what leaves the business, which is the strongest fraud and error control a small company has. Understanding why margin is what it is, since only the founder knows which delivery decisions caused it. And deciding concessions — discounts, write-offs, extended terms — because each one is a decision to earn less. Everything else is execution.

Yes, for two reasons. It makes the business fragile — an illness, a holiday or a bad week stops the month — and it makes the numbers unverifiable, because a figure only one person can explain cannot be checked by anyone. The fix is not another person memorising the same context. It is records where the explanation lives with the entry: evidence attached, reasons written down, corrections recorded rather than overwritten.

When the work becomes recurring and time-bound rather than occasional, and when its absence starts delaying decisions. Practical markers: invoices going out late during busy weeks, month-end requiring reconstruction, overdue customers going unchased until someone remembers, or a decision deferred because the position was unclear. None of these are about company size. They are about whether the numbers have become load-bearing while the person producing them has another full-time job.

It should mean the opposite. Handing over execution while disconnecting from the numbers is the worst of both arrangements — cost incurred, control lost. A founder who has outsourced properly looks at more, not less: a prepared cash position, the list of what needs approval, what is unresolved and who holds it, and monthly reporting with the decisions it raises. Less time producing figures, more time acting on them.

Three things, in a predictable order. The work moves to evenings and weekends, so it happens late and inconsistently. Invoicing and follow-up slip first, because they are the easiest to postpone, which puts cash under pressure. Then decisions start waiting for figures that nobody has had time to prepare. The damage is rarely a bookkeeping error — it is the compounding delay in everything that depended on the numbers being ready.