Our Cash Matching Reached Ninety Two Percent by Guessing Which Invoice Customers Meant

Description and use

Automated bank transfer matching reached 92 % using a fallback rule, but introduced errors. Removing the rule and adding manual review restored accuracy to 79 % automatic matches with under 1 % errors.

Problem

Automating the matching of bank transfer payments to invoices was difficult because payment references were inconsistent, leading to a low initial match rate of about seventy percent.

Outcome

The fallback rule temporarily boosted the automatic match rate to 92% but introduced a high error rate. Removing the rule and adding manual review restored accuracy, achieving a 79% automatic match rate with less than 1% wrong matches, improving credit control efficiency.

Workflow

• 1. Read bank statement each morning. 2. Search payment reference for invoice numbers. 3. Allocate payment to matching invoice. 4. If no reference matches, apply fallback: assign to oldest open invoice with same amount. 5. Audit fallback matches; identify incorrect allocations. 6. Remove fallback rule. 7. For unmatched payments, present candidate invoices to a person for manual review. 8. Report three metrics: automatic matches, manual matches, and wrong matches (sampled).

Evidence

At first the rate was about seventy percent, because references are messy.

Source

The most effective rule was a fallback. If no reference matched, take the oldest open invoice for that customer with exactly the same amount.

Source

When we audited a sample of fallback matches, about one in five had gone to the wrong invoice.

Source

The automatic rate is now seventy nine percent, and the wrong rate is below one percent.

Source