Data migration projects provide the complex data highway that enables information to travel from the old, legacy IT landscape, to the new, target IT systems.
Without a comprehensive data migration approach, any planned improvements for innovation, performance and growth, can be severely delayed, or worse, derailed completely.
Unfortunately, data migration is still poorly understood, particularly within management circles.
To help increase understanding and awareness of data migration, we recently commissioned a data migration research study with Data Migration Pro, the community resource for data migration practitioners.
Our goal was to understand the factors that lead to data migration success (and failure) so that those who are about to embark on a data migration can start to plan their approach by avoiding the many pitfalls that have beset projects in the past.
You can access the detailed 25-page Data Migration Research Study that outlines the key findings and a series of recommendations from the research.
I also caught up with the editor of Data Migration Pro, Dylan Jones, to get some highlights from the research.
Rebecca: What were the demographics of those in the study?
Dylan: It was a large study. We had 270 international respondents in total.
Although a real mix of practitioner roles took part, there was certainly a bias towards those with a Data Migration Lead background, which is great because obviously that role is key to any data migration.
Rebecca: Were there any findings that surprised you?
Dylan: I think the big one was the number of phased migration executions that are taking place now. We found that 62% of projects relied on phased migration execution.
However, looking at the research, it’s understandable when we consider that most migrations involved three or more legacy systems and of course data volumes are much higher these days.
A phased strategy requires much more rigorous legacy landscape analysis to figure out the implication of phasing your migration because you need to create a model that can cope with the transition of different data entities at different time-frames.
It’s also worth calling out that only 28% of projects experienced no data quality issues at all following their migrations. There seems to be the perception that poor data quality in the target environment is “the norm” for migration but with the right strategy and technology in place this is totally avoidable and can be planned for up front.
Rebecca: On the topic of challenges, was there anything the research uncovered that was a cause for concern?
Dylan: Yes, there were lots of areas for improvement.
In particular, many people cited poor scoping, planning and forecasting as a key challenge, which are all critical activities at the outset of course. Get these wrong, and the whole project can spin out of control. Only 46% of respondents felt their project was planned and forecasted effectively. 52% felt that the scoping carried out was effective.
So lots of improvement required there within the industry.
I think the lack of effective management and governance of data migration projects is also problematic.
We found that only 43% of projects had effective governance and management which is clearly a concern. I think this is symptomatic of a broader lack of understanding in general about data migration and what’s involved.
For example, we found that only 43% of projects had a project management team who had a good understanding of data migration best-practice, that’s much too low.
One of the Data Migration Leads we interviewed for the study vented their frustration that, even after buying everyone copies of best-practice guidebooks on data migration, the client still didn’t comprehend that this was a complex, challenging initiative that requires a lot of engagement and support from the business.
On the topic of engagement, we found the lack of business engagement a concern. We discovered that only 50% of projects had effective business engagement which is far too low given the significant role business users have on a data migration.
For example, in my past projects, I’ve had business users constructing data quality rules, setting up the user acceptance tests, collaborating on mapping, helping to document the system retirement policies – they perform a critical role so as an industry, we need to adopt better practices to get the business actively bought in and collaborating on our data migration projects.
Rebecca: What is one tactic people can draw from the research that is simple to execute?
Dylan: Good question. I’d say the big one that stood out is the need to always perform an impact assessment right at the start of the project.
So many of the problems we saw (e.g. poor planning, forecasting and scoping) were caused by a lack of thorough impact analysis from the start.
I’ve been preaching the benefit of a pre-migration impact assessment for many years now but the message still isn’t getting through to suppliers, customers and practitioners that you need to understand the scope and scale of the migration, including any data quality obstacles, before you even commit to the migration, ideally even before the contract is signed.
By performing a robust impact assessment, you can massively de-risk the project by equipping everyone with the accurate, timely intelligence on scope, data quality, relationships, stakeholders, and so much other information, that is so often missing from those early planning and initiation phases.