Reconciliation software and the bookkeeper: what it changes and what it cannot

Reconciliation software is having a moment, and the market-research headlines make it sound like the month-end close is about to run itself. It is not. What these tools actually do is take the mechanical half of reconciliation off your plate: pulling bank feeds, matching transactions in bulk, flagging the exceptions, and keeping an audit trail of who cleared what and when. The judgement half, deciding whether a mismatch is a timing difference, a missing entry, or a genuine error, still lands on the bookkeeper. This article is about where that line sits and how to use the software without letting it quietly become the reconciler.

What reconciliation software actually does

At its core, the software matches two lists. On one side is your ledger balance for an account, with its transactions. On the other is the external source: a bank statement, a credit card feed, a subledger, an intercompany counterpart. The tool compares them and sorts every item into matched or unmatched.

The value is in the matched pile. Bulk-matching hundreds of clean, one-to-one items by hand is the least valuable thing a bookkeeper does, and it is exactly what software does best. Rules can handle recurring pairs: the merchant fee that always posts two days late, the payroll transfer that always matches to the cent. Once those are automated, what is left on your screen is the short list that needs a human.

Where the judgement stays with you

The unmatched items are the reconciliation, and no tool can decide them for you. A deposit in transit is fine. A duplicate supplier payment is not. A cheque written last month that has still not cleared might be stale, or might be lost, and the write-back treatment depends on your jurisdiction and the client's policy, so confirm both rather than assuming.

The risk with automation is complacency dressed up as efficiency. If the tool auto-matches on fuzzy rules and you approve in bulk without reading, a wrong match clears silently and the account "reconciles" while being wrong. Our rule of thumb: automate the matching, never automate the sign-off. Review the exception list line by line, every month, and keep notes on what each item was and how it was resolved.

What the software does to the close

A good tool shortens the close by making reconciliation continuous rather than a month-end scramble. Feeds refresh daily, so by close day most items are already matched and the outstanding list is short. That is a real improvement to sequence: you spend close week on accruals, cut-off and review instead of on tick-and-tie.

But the close still needs the reconciling-items schedule. Software will show you what is outstanding; it will not tell you that a reconciling item has sat there for four months and is probably not a timing difference at all. Ageing the exceptions is still your job, and an item that never clears is often the first visible symptom of a deeper ledger problem.

Evidence and audit trail

One genuine advantage over spreadsheets is the built-in evidence. Who matched what, when, under which rule, with what supporting document attached: that is exactly what an auditor or reviewer wants, and it survives long after a working-paper spreadsheet has been emailed around and versioned into ambiguity.

Two practices keep that value intact. First, do not export to Excel to "finish" the reconciliation, because the audit trail stops at the export. Second, keep the sign-off inside the tool, with a reviewer distinct from the preparer, so the trail shows genuine review rather than one person clearing their own work.

When software is not the answer

If the ledger itself is damaged, duplicated, or carrying corrupted opening balances, reconciliation software will faithfully match garbage to garbage and report a clean result. The tool compares lists; it cannot judge whether either list is trustworthy. When an account will not reconcile because the underlying data is broken, that is a data-repair problem, not a matching problem, and it needs a specialist before anyone automates anything. The same applies after a botched migration: reconcile the opening balances by hand first, then let the software take over the ongoing months.

Choosing sensibly

Match the tool to the accounts you actually reconcile. High-volume bank and card accounts benefit most. Low-volume balance sheet accounts, prepayments, accruals, director's loans, may be faster in a working paper than in a tool built for feed matching. Whatever you choose, the test is the same: does it make the exceptions clearer, and does it preserve the evidence? If it does both, it earns its place in the close.