You Can Lead the Data Governance Horse to Water, but …
Data governance is a touchy subject. Many organizations have tried to build a practice…and failed. Other organizations feel too daunted to even try. IT in particular has a very conflicted relationship with data governance.
When IT management considers data governance, they want to be proactive, want to address issues, want to "shake the tree and make things happen." Being proactive is a generally a very good thing. But as it relates to data governance, the activities chosen by IT management often can be less than productive. Sometimes IT will take on the full burden so that data governance teams will be operated only by IT staff. Consequently, people on the business side never really engage as they fail to see a reason for being involved. At some point IT stops seeing a need to engage the business at all, thereby causing business to consider IT an out-of-touch problem for getting things done. And any case for the value of metadata may have been compromised by IT’s creation of wikis that contain virtually random bits of information that is consistently proven to be out-of-date or otherwise insufficient.
Other scenarios go in an opposite direction: IT orders the rest of business around, and makes them “do” data governance. When forced into such a position, the business leaders involved may not understand what is needed of them, and may not even want the added responsibilities. Under these kinds of circumstances things eventually devolve into a chaotic grouping of fits and starts that never progress. Data governance is existing in name-only and is simply circling-the-drain. When such strong-arm approaches don’t take, IT then blames the business, gets mad, and sulks. Or worse, the data governance failure is labeled a success and forever thereafter ignored.
Alternatively, IT can be proactive, but in a manner that is also playing a waiting game. IT can actively establish minimums of good data administration. IT can start putting the pieces in place that someday, down the road, will work towards enforcing business-led policy. This could mean having projects create process overviews, data dictionaries, various artifacts that start providing a picture of the data and processes by which the business operates. It means establishing a change management structure, establishing reviews to ensure everyone inside IT is following agreed upon practices. However, things are held back a little: perhaps each project has good documentation, but all projects are not organized together for an enterprise view; or data dictionaries exist but only the technical details are filled in, business details are filled in with “TBD.” Maybe a single administrator works to keep things together, when in fact, more resources are obviously necessary. The goal is to establish a framework of good practices to bring the IT-side of things into what might best be considered a “data governance ready” state.
While working at being data-governance-ready, efforts should be forwarded to expose business, as much as possible, to the documentation that has been put together. In these efforts, IT management needs to channel their inner-pied-piper to attempt to spark key business leaders’ interests in better organizing this metadata, seeing the benefits of business directed rules overseeing the work done in IT. Ideally, business will come quickly to understand the data as being used to create better insights, as well as in the value of data as that true corporate asset. And when business leadership is interested, maybe even hungry, to participate and help move the ball forward, then IT management can help place business leaders into the data governance leadership roles they need to assume, and transition into getting actual data governance in place and enforced. In a best case scenario, only minimal changes will be necessary in order to jumpstart into a maturing data governance practice.