Nov 2015 | Data Quality

What makes a successful data migration?

As a regular at Iergo’s Data Migration Matters event and host of our own migration themed roundtables (MMRT), I’m frequently asked for advice about how to manage common challenges experienced by those going through the process.

So today’s question to you data migrationeers is: How do I know when I have completed my data migration and how do I know I have completed it well? This inevitably always comes back to whether you can answer the question: How well did my data migration start?

The beginning is a very good place to start! One of the key discussion topics emerging from MMRT was: How do I start my data migration project in order to ensure the most successful outcome, and the most successful outcome first time? First we will define a data migration for the purposes of this discussion – a data migration is the “traditional” extracting of data from one (or more) system(s) and moving it to a different (new) system with the objectives, and this is key, of (i) transferring the day to day processing of the migrated data to the different system and (ii) not doing that same day to day processing on the existing system. This means that all or part of the existing system is, for all intents and purposes, being switched off and the data being migrated will not be available in the existing system – if parallel running is part of the project plan then the migration is not deemed to be completed until this phase is over.

If the above paragraph describes data migrations as you understand them, and there are many variations, then what does the switching off of an existing system imply? It implies that come the day of the migration that the business is happy that the existing system can be switched off, which in turn implies that we can identify the person in the business that is ultimately going to experience that joy – we have found our data migration business sponsor.

While we would expect our business sponsor to take responsibility for signing off the switch off, we would expect them to enlist a team of people to support that decision. Who constitutes this team and what their responsibilities are, is the subject of a future blog.

The motto of this blog instalment is: if you can identify your business sponsor you have a key ingredient for a successful beginning. If you cannot, then don’t take on the role of Data Migration Project Manager!

Some useful resources on this topic include our whitepaper, “Tackling the ticking time bomb: Data Migration and the hidden risks” by Iergo’s Johny Morris. You can also read more about Experian Pandora’s data migration capabilities in this area if you’re interested in technology to support your migration.

And my final thought… Defined and accepted data governance within the enterprise will strongly inform the identity of the business sponsor – if it doesn’t, then your data governance is probably not as strong or as effective as you thought it was. If that sounds like you then I’d recommend reading our whitepaper “Getting ahead of the game: proactive data governance”¬†for some practical advice on implementation.