“Data Governance” or “data governance”
There are many, many ways to tackle your data governance needs. Very often organizations start small and work their way out. Small successes are supposed to pave the way for larger ones. But at some point, the little iterations need to take on the enterprise role. If not, you get stuck in the realm of the lower case “g.” You never get to the big “G.” The worst case is you get multiple realms of governance.
Why is that? Why can’t we just roll up all the lower-case data governance work and call it some sort of gestalt DG? The answer is that not all DG capabilities are suited for local implementation.
Some capabilities under the DG umbrella can be localized or federated. Metadata, data controls, data integrity and privacy are good examples. Other capabilities are not remotely close to being suitable for a local, or department or project implementation. Thinks about DG capabilities like business alignment, data value measurement, or managing changes related to data management. These are enterprise level in context. These are upper case “G”.
I always hear the same questions at this point, so here are brief answers to two common questions:
How do you avoid getting stuck in localized capabilities?
- Do you want to get unstuck? Many times, cultures are impenetrable, and DG efforts cannot get any traction outside of specific projects. If this is the case, patiently pursue elevation of data literacy. Until leadership gets a new mindset, you are wasting your time.
- Assuming receptiveness to leveraging DG capabilities that span enterprise or are less federated, you need to look for opportunities to apply the new capability where there is a mechanism in existence, such as a PMO. A PMO is a great way to roll out new DG capabilities. (For a check list of PMO / DG capabilities contact John)
- If there is not a PMO, you need a strong sponsor. And better yet, a project that spans many areas such as MDM. Much the same way as a local roll out, you roll out the broader DG capability with the MDM program
How do you handle the transition from local to enterprise?
- Once you are recognized as a legitimate set of capabilities, you still need to implement them. For data people this can be challenging as they rarely get center stage and are at times ill equipped for the task. So, the initial activities must be around ensuring the transition team is well -sponsored and able to communicate well and operate across functional areas.
- Make sure that all stakeholders understand what the future looks like. What will be different?
- Design any new policies, architecture, standards, etc. etc. and vet them thoroughly. Start to train key stakeholders and make adjustments to your material
- Roll out the new capability, and measure everything you can think of!
Remember, while incremental implementation of DG is the preferred approach, at some point you need to make the “g” in “governance” a “G”. You need to transition your little “g’ increments to an audience who is ready for upper case “G.”