Manual journal reliance: when the ledger is being forced

By Peter Edwards ACMA CGMA · Chartered Management Accountant in practice

A healthy ledger is mostly written by documents. Invoices, bills, bank lines — each entry arrives with something behind it, and if a number is questioned, the evidence is one click away. A manual journal is different in kind: it's an assertion. Someone typed debits and credits because they believed the books should say something they didn't. Sometimes that belief is right and routine. But when journals become a large share of how the ledger gets written, the books have stopped recording the business and started being steered toward a result — and every journal is a door an error can walk through with nothing behind it.

Measuring it — one export, one column

The Account Transactions (or general ledger) export for the last year has a source column. Filter it to manual journals and look at two numbers: the count of journal lines, and their value as a share of revenue. There's no universal threshold, but the shape of the answer is usually unambiguous: a business whose journals are a small single-digit share of revenue, posted mostly at month-ends, is being maintained; one whose journals run at a quarter or more of revenue, scattered through the month, is being corrected — continuously, by hand, by someone compensating for a process that doesn't work.

The journals that are fine

Not all journals are confessions. Payroll postings from a payroll system, depreciation, accrual and prepayment releases, closing-stock movements — these are the ledger's legitimate month-end machinery: predictable, recurring, each traceable to a schedule that exists outside the journal. The test is whether the journal has a document or schedule behind it. A journal that references a payroll report is bookkeeping; a journal whose narrative is "correction" or "adjustment" or nothing at all is a symptom.

What heavy reliance is usually compensating for

• A bank feed nobody reconciles — so differences get journalled away instead of matched.
• Miscoding at source — costs landing in wrong accounts all month, swept into place by one big month-end journal that hides where they actually went.
• Inter-company and director movements with no agreed process, re-balanced by hand each period.
• A predecessor's habits — recurring journals copied forward for years after anyone remembered why.

Reviewing and unwinding it

1. List every recurring journal and write down, for each, what it compensates for. If nobody can say, that's the finding.
2. Trace the one-off "correction" journals for the last few months to what they corrected. Patterns will appear — usually one or two broken inlets generating most of the volume.
3. Fix the inlet, retire the journal. A coding rule, a reconciliation routine, an agreed inter-company process — each one replaces a standing journal with a source-driven entry.
4. Keep the legitimate machinery, documented: the month-end journals that remain should each reference the schedule that supports them, so the trail holds up when someone else — an auditor, a lender, a successor — reads the file.

In a pre-engagement review, journal reliance is the finding that predicts the experience of the engagement better than any other: it tells you how much of the monthly work will be archaeology. Price accordingly.

Related: The suspense account that won't clear · How to price bookkeeping cleanup work