Audit the field that should be unique.
- Choose a CSV or compatible JSON export.
- Enter a unique-key field such as account_id, email, SKU, or invoice_id.
- Run the audit to see duplicate groups, row numbers, row count, columns, and missing-value counts.
Check an export for repeated IDs, emails, SKUs, account numbers, invoice numbers, or another field that should be unique. The same audit also counts missing values by column.
Whole-row duplicate checks miss cases where the same account or SKU appears twice with different status, amount, owner, or date values. Auditing a field that should be unique surfaces those collisions for review instead of silently choosing a winner.
If the same export needs the same repair every week, define the owner, cadence, evidence, exception rule, and escalation path around the reconciliation instead of relying on another manual rescue.
Does the tool delete duplicates?
The audit reports duplicate groups first. A cleaned CSV download is available from the tool after review, but you remain responsible for deciding whether the chosen key and cleanup are appropriate.
What if the key field is not actually unique?
The audit will show repeated groups. That is a signal to review the data model or choose a more appropriate identifier before using the field for reconciliation.
Structural data QA is not accounting, legal, compliance, audit, or source-of-truth determination. Review material discrepancies with the responsible person.