“Why I hate Data Governance”

Data Strategy Leader Louis de Roo hates cleaning up. In this blog he explains why cleaning up your data, and Data Governance, is actually extra important.

As a child, I loved playing with Lego. And certainly not always according to the building instructions; I often threw all the bricks into a pile, and then started a fantasy construction that was bigger than any regular Lego set. The downside was that this inevitably created some building rubble, and the growing pile of Lego bricks in the room didn't make it any easier to search for that one piece that had to fit my construction, so before long my building site took up half the living room floor. Just as inevitably, there was my mother, who firmly ordered me to clean up that mess before dinner.

When we work with data, we also build. New insights, decisions, and ways of thinking. And there too, mess often arises. Because not every source is equally reliable. And where Lego bricks still have a fairly consistent fit, that's already much more doubtful with some data. But during your journey of discovery toward a new data product, you inevitably run into the fact that not all data is equally good or reliable. Time to clean up the mess, then.

"Time to clean up the mess"

I don't always enjoy cleaning up. It not only forces you into an action you didn't ask for, it also forces a decision about what you do and don't find important; what you do and don't want to keep, and in many cases also a decision about what you still expect to need in the future. While cleaning up, I quickly get distracted by all the new possibilities I see while categorizing, and I'd rather look at what other constructions I can build. But there's my mother, still standing sternly in the room, and all the castles in the air disappear back into the storage box.

My mother was of course right. If I had never cleaned up all that Lego, my construction would have kept growing, but it would also have become increasingly difficult and impossible to keep the mess around it under control. You see this in many organizations too: as their data landscape grows in complexity, it becomes increasingly difficult to keep data quality and reliability under control.

Cleaning up and organizing

So it's important to start cleaning up and organizing before the task grows over your head. Preventing a mess is a lot easier than cleaning one up. Although it's not always a popular step; in the energy of the moment, nothing dampens your enthusiasm like being told you need to work in an organized way. Management and policy can quickly feel like a limitation, and that can be quite a dampener on your creative process.

But it is necessary: anyone who wants to keep trusting the quality of their data will occasionally need to ease off the enthusiasm a bit, and think carefully about the durability and reliability of that data. That may be a bit of a dampener, but it's wise. Data Governance is an important condition for not getting lost in your own mess in the long run. To ensure our constructions can grow with us and remain reliable and available in the long term. So that ultimately, new insights can also find a place in our data landscape.

Data Governance

So I hate Data Governance (a little bit). Because I love free association, creative space, and inspired exchange of ideas. Always tricky when a spoilsport like that comes around the corner. If you focus too heavily on Data Governance, you risk stifling innovation in the bud. It's important then to consider what level of Data Governance your organization needs: as much as necessary, but no more than needed. Is quality and reliability of the greatest importance for your operation? Then you'll need to set up your Data Governance much more robustly than if you're a company that mainly steers on coarser patterns in the data. Consider whether you work with confidential or sensitive data, and what the impact of incorrect data could be. How your organization handles data determines whether you'll need to do a lot or just a little with Data Governance. But for every organization, doing nothing is not an option.

In the end, you'll still have to clean up your room at some point!

Written by Louis de Roo
Data Strategy Leader
E-mergo