Delen via

Use partial matching in data reconciliation (preview)

  • Artikel

[This article is prerelease documentation and is subject to change.]

This article describes how to enable partial matching and use it while you reconcile data in Excel by using Microsoft 365 Copilot for Finance.

Partial matching extends the matching rules for data reconciliation. When partial matching is used alongside exact matching, the matching rules can identify not only exact matches but also similar values in your data. Specifically, Copilot for Finance uses exact substring matching on a selected pair of mapping keys. In this way, it can recognize occurrences of one string (substring) within another string (superset), exactly as they appear.

To help reduce the number of false positive classifications, all transactions that might potentially match are evaluated on a one-on-one basis.

Apply partial matching for a mapping key

When you define reconciliation vectors, you can indicate a pair of columns that you want to apply partial matching to.

To apply partial matching for a mapping key, follow these steps.

  1. Open an Excel workbook that contains the data that you want to compare.
  2. Follow the regular process of data reconciliation until the step where the vectors are presented. Learn how to reconcile data in Reconcile data with Copilot for Finance.
  3. Select Keep to start to edit the mapping keys.
  4. Hover over the monetary key where the tolerance is to be applied.
  5. Select Add modifier.
  6. Select Partial matching.
  7. In the Reference column drop-down, choose the column to use as the superset.
  8. Select Save
  9. The partial matching label now appears for the selected mapping key and the superset column contains the "REF" (reference) label.

Note

For more information about the reconciliation process, see Reconcile data with Copilot for Finance.

Data classification for partially matching transactions

All transactions that are identified as partially matching are classified into two groups:

  • Unmatched transactions

    • Values in the selected mapping key partially match, but there's a discrepancy in the monetary values.
    • The transaction doesn't have a corresponding match in the other table.

  • Potentially matched transactions

    • Values in the selected mapping key potentially match, and the monetary values are equal.

There's no case where partially matching transactions are classified in a matched transactions (perfectly matched) category.