Skip to content
Ixia Consulting

Migration Assurance

Getting Data Loaded Is Not the Same as Getting Data Migrated

A load that ends without errors proves the tooling worked. It does not prove the business received its data.

Dave Welensky, Director · Published August 2026 · 5 min read

The most dangerous moment in a migration is the successful load, because success is declared in technical terms while the risk is carried in business terms. Ixia's migration methodology was built around that distinction, and it is the difference between loaded and migrated.

The load that worked

Every migration has this meeting. The load finished overnight, the error log is short, the record counts match, and someone proposes telling the steering committee that data is green.

In migration programmes, one of the first things we look for is what green actually means. A successful load tells us the extract ran, the transformations executed and the target accepted the records. It does not tell us the business can invoice, pay, pick or report. Those are different claims, and they need different evidence.

Four gaps that hide inside success

Extraction is not migration. Pulling data out of a legacy system proves access, not meaning. The mapping decisions that turn a legacy field into a target field are where migrations are won, and none of them are visible in a load log.

Load success is not business success. The target accepted the record because it was technically valid. Whether the account group, payment terms and organisational assignments are right for how the business will actually use that record is a question the load never asked.

Record counts are not reconciliation. Counts reconcile long before values do. A migration can land every record and still move the debtors book by millions, because reconciliation lives at the level of amounts, balances and relationships, not rows.

Technical validation is not business validation. The tooling checks what the tooling can see: formats, mandatory fields, referential integrity. Whether the migrated stock levels let Monday's replenishment run correctly is validated by the business, on evidence designed for business eyes.

What migrated actually means

Over many programmes Ixia developed and applied its own data migration methodology, IDMM, precisely because the standard definition of done stopped at the load. Migrated, in our definition, means the data has been profiled, mapped against business meaning, transformed, cleansed, validated, loaded, reconciled to value level, and signed off by named business owners on evidence they could read.

It also means repeatable. A migration proven once in a mock cycle and executed identically at cutover is engineering. A migration assembled by hand on the final weekend is gambling with a calendar.

A successful load tells us the system accepted the data. Sign-off tells us the business accepted it. Those are different events, usually weeks apart.

Dave Welensky, Director, Ixia Consulting

The question to ask your programme

If your migration status is green, ask what the green is made of. If the answer is load statistics, you know the loading status. The migration status is the reconciliation evidence, the exception queue and the list of business owners who have signed. When those are green, the weekend is just the performance.

Key takeaways

  • Load success is a technical claim. Migration success is a business claim with evidence.
  • Counts reconcile long before values do. Reconcile values.
  • Business validation needs evidence designed for business readers.
  • Repeatability separates engineering from luck at cutover.

Recognise your programme in this?

Written by the people who deliver the work. The conversation works the same way.

Discuss migration assurance