The problem is that many project leaders simply have no experience of delivering a data migration. It’s often assumed that it’s “just another IT project” (it isn’t) and the focus is on shunting data around so it’s the responsibility of the IT team (it isn’t).
Coupled with this is the challenge posed by the “chuck it all over the fence to the systems integrator” tactic that many companies choose to adopt in a hope to de-risk and accelerate their data migration requirements.
With the wrong systems integrator this amount of devolvement can easily get out of hand as the control of the project slips away from you.
So data migration leaders continue to ask themselves: “Why is this thing so damn hard?” and recent research still points to an unacceptably high failure rate of data migration projects.
It doesn’t need to be this way.
There are some simple “pillar” tactics that data migration leaders should adhere to. If you get these right you stand a far higher chance of migrating data successfully and creating a positive experience when the target system goes live.
Here are some of my favourite leadership tactics for data migration but feel free to add your own:
Pitfall #1: Ignoring the Complexity of Deliverables Creation and Coordination
Don’t wait for the project to start before figuring out what documents like a “system retirement plan” or “data dictionary” will look like.
It’s a common pitfall for project leaders to find themselves struggling with the complexity of the many types of resource information, specifications and plans that get created right from the outset to the very end.
I urge you to figure out exactly what you’ll need before the project kicks off. Understand how all the deliverables relate to each other and who should complete which sections.
It’s a simple pitfall to fall into but one that can really hamper your results as a data migration leader.
Pitfall #2: Leaving Data Quality Strategy to Chance
Data quality management is the bedrock of your data migration.
Ignore it and you’ll watch your budget dwindle as the testing team mop-up post migration defects. Get it right and you’ll see your project teams humming along with ease, buoyed by the certainty that the data they’re working with is well managed and coordinated.
There is nothing more damaging to the success of a data migration project than to leave data quality to guesswork and assumption.
For example, many organisations are genuinely surprised when their contractor pushes back data to be corrected by the business. Assumptions and “gentlemen’s agreements” over the responsibility lines surrounding data quality need to be replaced with solid contractual statements and written agreements.
As a data migration leader you have to be crystal clear about what framework is required for effective data quality before, during and after the data migration.
Pitfall #3: Creating Inaccurate Project Forecasting and Resourcing Estimates
You can’t measure what you don’t know so how can you measure the duration, cost, skills and danger areas of your data migration without simulating what will happen months down the line?
The answer is you can’t which is why you definitely need to push for a pre-migration impact assessment so that you can discover project pitfalls in advance of the project initiation phase.
This is such a straightforward tactic to adopt but very, very few projects implement it. I’ve implemented it numerous times, from multi-million projects all the way through to small businesses, and every time it’s more than returned the investment.
It does require some specialist skills and technology (if you don’t have them already) to perform a rapid pre-migration impact assessment but again, the costs of leveraging these for a few days to gather insights is well worth the investment. I’ll be explaining this tactic in more detail over the coming weeks.
Next Steps
This is my first post on the new Insights series and I’ll be providing more practical techniques to combat common Data Migration pitfalls over the coming weeks.
Why not help me by posting some of the most common pitfalls you’ve witnessed in the comments below so I can try to tackle them one-by-one?