I attended a breakfast discussion on data governance this morning. I was given a copy of the book “Data Governance For The Executive” by James Orr who was one of the presenters.
Here are a few brief — albeit disjoined — notes of what I found most interesting:
When looking at the evolution on IT systems from the mainframe days on, one can see an evolving perception that in the beginning code was deemed to be more important than data while nowadays data is clearly more important than code.
Studies show that the growth in unstructured data is not matched by a growth in the management of that data.
To ensure semantic interoperability, some organizations focus on the desired outcomes which in turn dictate the activities, which in turn define the resources necessary. Others start with defining a common business glossary.
A main question discussed was on how to infuse data quality into agile development. Data management needs to become critical of the mission of the organization and needs to be embedded into the software development processes, and not an afterthought. Without top down support and sufficient failure, any such effort is doomed to fail.
Another important question discussed was on what is core data; what data is important to key business processes, on a departmental level and what is critical to the enterprise. The flip-side is that if some data is not important we might not need to collect it.
Always start with why? What is the benefit? What is the compelling rationale for the initiative? Be mindful of the fact that business processes are tightly linked with the creation of data.
Tie operational incidents to data quality and data governance. Tie data governance to the prevention of exposure.
Crisis is a terrible thing to waste; a disaster – even more so.
Be mindful when to use tech language and when to use business language during the necessary translation between business and IT.