Docs
Browse Docs
Why clean data matters · Finance teams

A vendor list that paid the same invoice twice

An accounts payable team reconciling its vendor file against a year of invoices finds duplicate vendors, and duplicate payments behind them.

Illustrative example. The organization and its data are invented to show a typical pattern. This is not a customer story.

What they hold

  • A vendor master file with names, addresses, tax identifiers, and payment details.
  • Invoices entered by hand or read from scanned documents.
  • Payment records from the bank, and the general ledger.

The mess

One vendor set up more than once

Harlow Supply Co., Harlow Supply Company, and HARLOW SUPPLY each have a vendor record, created by different people at different times. The addresses differ only in how Street and Suite are abbreviated.

Invoice numbers that almost match

The same invoice is entered once as INV-00871 and again, from a reminder copy, as 871. The system checks for an exact duplicate invoice number per vendor record, and with two vendor records and two spellings it finds none.

Amounts in more than one notation

Credits appear as negative numbers, as amounts in parentheses, and as positive amounts with CR beside them. A total that reads the last kind as a charge is wrong by twice the credit.

Dates read in the wrong order

Invoices from an overseas supplier write the day first. Entered as written into a system expecting the month first, some due dates move by months and others are rejected.

What it costs

  • The same invoice is paid twice. Recovering the money depends on someone noticing and on the vendor's goodwill.
  • Spend with a vendor is split across its duplicate records, so the team negotiates without knowing how much it really buys.
  • Month-end close is delayed while totals that should match are reconciled by hand.
  • Early payment discounts are missed and late fees incurred because due dates are wrong.
  • An audit finds the duplicates before the team does.

What clean looks like

  • Each real vendor has one record, matched on tax identifier and confirmed by a person where the identifier is missing.
  • Invoice numbers are compared after removing prefixes and leading zeros, so near-identical numbers for one vendor are caught before payment.
  • Every amount is a signed number, with credits negative.
  • Dates are stored year first, and each supplier's convention is recorded once instead of guessed per invoice.
  • Every merge of vendor records is logged, so the history of who was paid, and as whom, survives.

Problems in this example

Karg checks for many of these when it profiles a file, applies the fixes it is confident about, and sets the rest aside for your review. See how it works, or have our team clean your data for you.