Risks and Failures of Data Migration

Data migration is high-risk because business operations depend on the data being accurate, available and secure. Even a small migration error can affect customers, employees, financial reporting, regulatory compliance and decision-making.

What failure looks like in practice

• Records missing after the move (row counts don't match)
• Key fields empty or truncated
• Dashboards that worked yesterday now throw errors or show wrong numbers
• Customer profiles incomplete
• Transaction histories missing or broken
• Downstream integrations failing silently

1. Data loss

Data that existed in the source never arrives in the target, or is destroyed along the way. Common causes:

• Records filtered out by a faulty transformation or a failed batch
• Truncation (a VARCHAR(50) target for 80-character values)
• Failed jobs that are not retried or reconciled
• Hardware failure, accidental deletion, or a malicious attack during the move
• No backup taken before cutover

Mitigation

• Take full, tested backups before migrating.
• Reconcile row counts and control totals (for example, the sum of invoice amounts) between source and target.
• Log rejected records and investigate them rather than dropping them.
• Keep the source read-only and available until sign-off.

2. Data corruption

An unintended change to data while it is written, read, stored, transmitted or processed. Common causes:

• Incorrect transformation logic
• Incompatible data types (decimals rounded, dates shifted by time zones)
• Character encoding problems (José becomes José)
• Software bugs in migration tools or scripts

Corrupted data is dangerous because it looks valid. A price of 19.99 stored as 20 passes every not-null check. Only thorough validation catches it.

Mitigation

• Compare checksums or hashes of rows or files.
• Validate at field level on samples and on critical columns.
• Test explicitly for encodings, time zones and precision.
• Have business users verify known records.

3. System downtime

The period when systems are unavailable during or after migration. Common causes: the cutover requires freezing the source, servers are overloaded by migration traffic, failures at go-live force troubleshooting, or a rollback is slow or untested. For e-commerce, banking or healthcare, even an hour of downtime means lost revenue and customer trust.

Mitigation

• Use phased (trickle) migration with change data capture (CDC) to keep downtime near zero.
• Schedule cutover in low-traffic windows.
• Run load tests and dress rehearsals.
• Define and test a rollback plan with a clear go/no-go decision point.

4. Security breaches

Unauthorised access to data, systems, networks or facilities. This is deliberate harm, unlike accidental loss or corruption. Attackers aim to cause financial or reputational damage through data theft, ransomware and leaked personal data.

Migrations increase exposure because:

• Data is copied to staging areas, exports and temporary files
• Data travels across networks and often out to the cloud
• Service accounts and credentials are granted broad access
• Access controls may not be recreated correctly in the target

Mitigation

• Encrypt data in transit (TLS) and at rest.
• Use least-privilege, time-limited credentials for migration jobs.
• Mask or tokenise sensitive data in non-production environments.
• Securely delete staging copies and exports after the migration.
• Keep audit logging and monitoring on throughout.

5. Compliance violations

Breaking legal or regulatory obligations during or because of the migration. Examples:

• Moving EU personal data to a region outside the EU in breach of GDPR data residency rules
• Losing the audit trail on financial records (SOX)
• Exposing patient data (HIPAA) or card data (PCI DSS)
• Failing to carry over retention and deletion policies

Consequences include heavy fines, failed audits, legal action and reputational damage.

Mitigation

• Involve compliance and legal teams from the planning stage.
• Classify data (PII, financial, health) before moving it.
• Document data lineage and keep migration audit records.

Accidental versus deliberate risk

Why migrations fail

• Underestimated complexity and poor source data quality
• Inadequate testing and validation
• No rollback plan
• Weak involvement from the business and data owners
• Unrealistic timelines

Key takeaways

• Small errors have large downstream effects on customers, finance and compliance.
• The five core risks are data loss, corruption, downtime, security breaches and compliance violations.
• Corruption is the hardest to detect, so validate thoroughly.
• Security breaches are deliberate and need dedicated controls.

Next: the business side, identifying requirements and building processes for a smooth migration.