Data migration is not complete because a technical load succeeds. A migration can move every expected row and still produce a system that the business cannot safely use. The assurance question is therefore not whether the process ran, but whether the resulting data is complete, correct, reconciled and operationally acceptable.
What data mapping actually proves
Source-to-target mapping defines how data should move and transform. It is essential, but it is only a design artefact. Governance still needs evidence that the mapping was implemented correctly, that transformation rules behaved as intended, that exceptions were controlled and that the migrated data retained the business meaning required in the target system.
Why row counts are weak assurance
Row counts can establish a basic level of completeness, but they cannot prove correctness. Values may be transformed incorrectly, reference data may be misapplied, relationships may break, duplicates may be introduced and exceptions may be silently excluded while the headline count still reconciles.
Evidence required for migration confidence
- approved source-to-target mappings and transformation rules;
- data profiling and known quality issues;
- reconciliation at record, control-total and business-value level;
- exception reports with ownership and disposition;
- repeatable migration rehearsal evidence;
- business validation of critical data domains;
- interface and downstream dependency checks;
- clear acceptance criteria and accountable sign-off;
- cutover sequencing that recognises migration dependencies and recovery options.
The client-side control
The supplier may own migration tooling and execution, but the client owns the consequence of incorrect live data. Independent assurance creates the necessary separation between the team performing the migration and the judgement about whether its evidence is sufficient to support progression or go-live.
Enigma applies this principle through its Evidence-to-Decision Assurance Model, where migration evidence is assessed for relevance, completeness, traceability and fitness for the decision being made.
The decision standard
A successful migration run proves that a process executed. It does not prove that the organisation should trust the data. Readiness exists only when evidence demonstrates that material data is present, correct, reconciled, exceptions are understood and accountable business owners have accepted the result.