Lucknow, India

Data Management

Prepare a data migration reconciliation checklist

Compare source and destination meaning, relationships and exceptions before approving a migration.

IGS Editorial · 3 min read · Updated 17 November 2026

Prepare a data migration reconciliation checklist — editorial artwork
Illustrative data management image

A data migration is not complete merely because the destination contains the expected number of rows. Values can be transformed incorrectly, relationships can break and important records can be excluded while the total still looks plausible.

Reconciliation connects the source, transformation rules and destination. Plan it before migration so the required evidence is available. If the source changes during the process, define how those changes are captured and compared.

Inventory what is moving

List entities, fields, attachments and relationships in scope. Record exclusions and the reason for each. A migration may intentionally leave historical material behind, but that decision should be approved and understood by users.

Identify stable keys for comparison. Names or row positions are often unreliable. Where identifiers change, maintain a mapping that supports validation and later investigation.

Validate transformation rules

Document field mappings, default values, format conversions and lookup changes. Test ordinary, empty and unusual values. A transformation that works for one date format may misinterpret another source's records.

Check whether meaning changes. Combining several status values into one destination category may lose information users still need. Have the business owner approve that consequence rather than treating it as a purely technical mapping.

Compare at several levels

Use record counts, key totals and field-level checks where appropriate. Check relationships such as orders linked to customers or documents linked to cases. Totals alone can conceal records assigned to the wrong entity.

Select representative records for end-to-end review in the destination interface. Users should be able to perform the intended task with the migrated information, not just locate a database row.

Account for exceptions

Record rejected, changed and unmatched records with reasons. Assign an owner to decide whether each should be corrected, excluded or handled through another route. Do not quietly discard failures to make the migration report appear complete.

For a missing reference, establish whether the source is incomplete or the mapping failed. The resolution may require a business decision. Preserve the evidence used so later reviewers can understand the result.

Define cutover and recovery

Agree when the source stops accepting changes, how final updates are captured and who approves the destination for use. Identify a recovery plan appropriate to the system and test its practical availability.

Keep reconciliation results, approvals and unresolved limitations with the migration record. After launch, monitor tasks that depend on the migrated data, because some issues emerge only in use. A careful reconciliation plan makes the transition explainable and gives users a route to report discrepancies without losing their connection to the source.

Reconcile a relationship, not just a total

A migration transfers the expected number of orders and customers, but some orders point to the wrong customer after identifier conversion. Row totals alone will not detect that problem. Validate the relationship using stable mappings and representative records.

Check how missing references are handled. A default placeholder may preserve the row count while making the record unusable. Keep rejected or unmatched relationships visible and assign an owner to resolve them.

Have a business user perform a realistic task in the destination, such as finding the correct order history from an authorised customer view. Compare it with the approved source evidence. Record any discrepancy and the rule or mapping responsible. Before cutover, decide which unresolved limitations are acceptable and which block use. This ensures the approval reflects practical usability and data meaning, not only technical completion. Preserve reconciliation evidence so later questions can be connected to the migration decision and source version.