When a firm neglects its own books

Not a competence problem

Nobody running a bookkeeping practice has forgotten how to reconcile an account. Firm books drift for a duller reason: client work has an external deadline, a person waiting, and an invoice behind it. The firm’s own close has none of the three, so it goes last, and last is a position you can occupy forever.

The excuses are consistent enough to be a diagnostic.

What gets said	What is happening
“My team handles it”	The team handles client files. The firm file sits below all of them in the same queue
“I’ll do it in the slow season”	There is a busy season and a recovering-from-busy-season
“It’s not that messy”	Twelve to eighteen months of drift is where “a bit behind” becomes reconstruction. Nobody notices crossing the line
“I check the balance”	A bank balance is one number on one day and tells you nothing about margin

Count hours, not dollars

Four hours a month on the firm’s own books, plus an hour of owner time reviewing what those four hours produced, is 60 hours a year. Those hours come out of the same pool as billable work, so the real cost is not the labor. It is the file you did not take on.

The compounding version is worse and less obvious. A firm file left nine months does not need nine months of work. It needs nine months of reconstruction, which runs slower per month than a current close, because the source records have to be gathered before anything can be entered. You already sell this asymmetry to clients. It applies at home identically.

The firm file is genuinely harder than a client file

This is the part that catches experienced people out, and it is why competence offers no protection. Practice books carry transaction types most small-business files never do.

Owner draws and partner distributions. Equity movements, not expenses, and the single most common miscoding in a practice’s own file. They are hard to catch because bank rules code by payee, and a transfer to a personal account looks like every other transfer.

Quarterly estimated payments. Frequently made from a personal account and never recorded, which shows up as a reconciliation difference with no obvious cause and gets written off as “something odd in Q2”.

Software, billed three different ways. Annual, monthly, and per-seat, often across two cards, often with one vendor appearing under three payee strings. Your subscription spend is wrong and no total will tell you so. You have to open the payee list and look.

Contractor payments. A missing W-9 in January is admin. The same missing W-9 in December is a problem, and if you pay subcontractors this is the item most worth automating.

Equipment bought outright versus financed. One is a fixed asset with a depreciation schedule. The other adds a liability and needs the payment split. Expensing either is an amended return.

None of that is difficult. All of it is easy to defer, and every item is invisible in a bank balance.

What does a firm’s own month-end close contain?

Discipline is not the fix. A named job, on a fixed date, owned by someone who is not the person furthest behind on client work, is the fix. In order:

1. Reconcile every account to the statement. Operating, savings, each card. To the statement, not the feed.
2. Clear uncategorised to zero. Not “mostly”. Anything you cannot identify goes to a suspense account with a note attached, which is a question to answer next month instead of a place to hide it.
3. Post draws and distributions to equity, and check nothing landed in expenses by rule.
4. Record the quarter’s estimated payment in the month it was actually made, including the one paid from a personal account.
5. Open the payee list and look for duplicates. Same vendor, three spellings. Merge them.
6. Check contractors against W-9s on file. Monthly, so December is uneventful.
7. Read this month’s P&L beside last month’s. Not against budget. Two months side by side is where a miscoding shows up as a line that moved for no reason, and it takes about ninety seconds.

For a practice with one operating account and one card, that is under an hour once the file is current. Which is the entire argument for getting it current.

The part that shows up in how you sell

Your own books are the only work sample a prospect can reason about without becoming a client first. Not because anyone asks to see them, since nobody will, but because the questions they do ask get answered out of that file.

What does a file this size usually take you? What is your capacity next quarter? Do you make money on cleanup work, or does it just fill gaps?

A practice that answers those from data sounds different from one answering from memory, and firm owners can hear it immediately. Most of us have been on the other side of that call.

If it has already drifted

Months rather than weeks behind is a cleanup or a catch-up, priced the same way any file of that size is priced: per month of backlog, published up front.

$199 per month of backlog

The distinction is whether the months exist and are wrong (cleanup) or were never entered (catch-up). Most drifted firm files are some of each. The free 48-hour file review returns the split in writing and stays with you regardless of what follows.

If the work is going outside the practice, whether to subcontract at all is the decision that comes first, and what to check before you hand over a file is the one after it.

Not running a firm? The same problem, and the same fix, for a small business: bookkeeping for business owners, every fee published.