Duplicate payment detection

Duplicate payments occur when the same invoice is paid more than once, often because vendors submit invoices through multiple channels, invoice numbers vary slightly, or manual data entry introduces inconsistencies. The detection rate measures how effectively the accounts payable process catches these duplicates before money leaves the account.

Recovering duplicate payments after the fact is time-consuming and often unsuccessful. Vendors may be slow to issue credits, and the administrative cost of chasing recoveries can approach the value of smaller duplicates. Prevention through pre-payment detection is far more cost-effective than post-payment recovery.

Duplicate Payment Detection Rate = (Duplicates Caught Before Payment / Total Duplicates Identified) x 100

For example, if the AP team catches 45 duplicate invoices before payment and discovers 5 more after payment has been made, the detection rate is 90%. Track both the detection rate and the absolute value of duplicates prevented. A high detection rate on a large volume of duplicates may indicate upstream data quality problems that need fixing rather than better detection alone.

Implement automated matching rules that flag potential duplicates based on invoice number, amount, vendor, and date combinations. Use fuzzy matching to catch near-duplicates where invoice numbers differ by a digit or formatting varies. Establish a single point of invoice receipt so that submissions through multiple channels are consolidated before processing. Require three-way matching (purchase order, goods receipt, invoice) for all payments above a materiality threshold. Run periodic audits on historical payments to identify duplicates that slipped through and use the findings to refine detection rules.