In my last post, I introduced the “5 Keys” series, covering all of the core principles of enterprise information management (EIM). Today we introduce the first EIM discipline you should think about: data governance. The remaining EIM disciplines can be addressed in any order (as long as you keep balance), but data governance is a little different not just because it sits in the middle, but also because it appears slightly larger, as we see in the graphic below:
Figure 1: 10 Data Management Disciplines (adapted from the Data Management Association)
I assure you that graphic isn’t one of those optical illusions that make the circles appear different even though they are actually the same size. Hopefully by reviewing the following keys, you’ll understand why.
KEY 1: Data governance has a reputation for being overly burdensome, but it all comes down to letting people know when it is their turn to make a decision
This key is one that I literally run around my office telling people. Okay, I walk, but I do tell people in my office. “Data governance” is an unfortunate term that reminds you of dysfunctional Illinois governors, big brother watching, or anything but a worthwhile investment in your organization. With that word association and many data governance programs overly focusing on administration instead of business impact – it is no surprise that data governance is often viewed as a “nice-to-have” capability.
At its core, data governance is nothing more than a complex decision-support framework coupled with an organizational mandate and the corresponding decision-making authority. (See, I can sound like a real consultant when I try.)
What this means in human-speak is that data governance determines when decisions need to be made and then provides a way to get those decisions made. Sometimes this responsibility falls on data stewards, other times data owners, or sometimes it needs to be resolved by a collaborative executive team like a data governance council. Data governance sorts it all out!
KEY 2: The surrounding circles are like an orchestra with Data Governance as the conductor
This should be straightforward: data governance pulls the strings, and the other 9 EIM tools are the strings. To complicate it slightly, when an issue is identified by data governance the first step is to determine which of the EIM tools to use, and then the right person(s) to use them.
You can set it up with a matrix, with one axis having all the EIM tools, and the other axis having different people. It could be broken up by business units, with data owners, stewards, etc. This is typically one of the first activities you do in documenting new issues and preparing for prioritization and assignment.
KEY 3: Data Governance should not be another administrative hurdle – it should be an enablement engine
As a classically-trained programmer, I hate doing pointless documentation. I bet that you, and everybody you have ever met, feel the same on this issue. Be advised that data governance when done poorly can lead to enormous amounts of pointless documentation. I therefore advise not doing data governance poorly.
Data governance is about enablement – like we talked about above, it is about letting people know when it is their turn to make a decision. When you look at your workflows, evaluate them through that lens – is this contributing value to decision-making or is it pointless and should be eliminated? Get rid of all those that aren’t contributing.
We should judge more of our business processes through simple mechanisms like this!
KEY 4: Data Governance can’t govern without the other 9 EIM tools, and the other 9 EIM tools can’t do much without Data Governance
I talked about balance in the introduction to this “5 Keys” series. This key takes the concept of balance to an almost-paradoxical level. The trick is that the other 9 individual EIM tools can’t do “much” without data governance – they can still do a little!
One of the most effective ways to get started is with metadata management. Do some basic business metadata capture – maybe a set of 20 data elements and their business definitions. Just capture it all in an Excel spreadsheet and then validate the definitions with people you think should be involved with your data governance. These simple actions will plant the seeds of metadata management and data governance, and wondrous lush EIM will grow before you know it!
KEY 5: Whether or not you do Data Governance, you have Data Governance
If your organization has data and does something with it, you have data governance. It may be that you are entrusting all of your data quality, security, data development, and all the other EIM responsibilities to the developer in the corner – but somewhere, somehow those functions are being performed. Having been that developer in the corner, I can assure you: they do not always perform these functions with the same rigor or consistency you might prefer.
This key may be the most important of any in all of this “5 Keys” series. EIM is about using these 10 tools to take deliberate actions to manage your important data assets. It is about shining light on the processes that otherwise are happening uncontrolled in the shadows of your organization.
And before you run off to blame those developers, consider this analogy:
The data assets in your organization are a crucially important ingredient – like water. Your unmanaged, in-the-shadows data governance has built you a bucket to hold the water, but it leaks and you can never find the bucket when you need it. By adopting deliberate EIM, you will build indoor plumbing – with water that comes out of the faucet whenever you simply turn the knob.
So the next time your organization considers data governance a “nice-to-have”– it absolutely is. You can always take the bucket down to the well – unless of course, that bucket is missing again.
Anthony J. Algmin is a Manager in the Business Intelligence Practice at West Monroe Partners.