Bringing Data Governance Into Focus
Posted by 
Share this Post
Marty explains how to bring your data governance into focus
In last week's blog I attempted to define what I mean by "master data" and "reference data." I also found some good material by my colleague Larry Dubov on the topic. In this installment, I want to address how you pick a good subject for a data governance increment.
Assuming that you can't solve all data quality issues at once, much less even identify them all, the question arises: how do you decide which set of issues to focus on? Let’s discuss that today.
Imagine you were to visit the top leaders in your organization today, and asked them for 30 minutes of their time to discuss the "things that keep them up at night" for the purpose of prioritizing the "biggest bang for the buck" projects to start making those things go away.
In such a meeting, it would be fairly easy to get a list of 10-12 issues, risks, concerns, conflicts, costs and missed opportunities that bug your executives. If you met with a dozen of these top executives, you would quickly develop a list with a lot of common themes. These are the topics that matter to your executives.
If you then asked them to prioritize the list, you'd be surprised at the consensus of the "top" issues. Now, you're probably muttering to yourself, "This is a gross oversimplification," and you're correct.
My point is that you don't need to gather these people for a two-day offsite to determine this. But through this exercise you'll discover which execs are passionate about eliminating business issues by fixing data quality problems. This will be helpful later when you go back and ask for their help!
This is the list of pains and opportunities that you need to start with. Next, you need to examine the root causes for these issues - are they caused by process or workflow inefficiencies or by poor quality data, or some combination of these?
In most commercial organizations, whether B2B or B2C, the company has issues with customer and/or account and/or contact data. Most every business leader will be affected in some way by poor-quality data as well as some broken processes.
This analysis is somewhat difficult if you're looking for hard, irrefutable evidence. But at this point, all you want is a list of issues for which data quality is a key culprit. This list gives you the starting point you can revisit with your executives to get their feedback.
With this matrix of business issues tied to data subject areas and high level business processes that are "broken," you can perform a simple affinity analysis to see which are common data and process issues.
You'll likely see natural groupings of these, and there will likely be several groupings that appear. This is what you'll use to meet with these business leaders in a group setting to establish priorities for which problems could/should be solved first, second and third. You’ve likely already piqued these leaders’ interest with your questions, so capitalize on that.
If you've followed Larry's recent blogs about prioritization for MDM projects, you'll find more explicit explanations of this, but the overall approach is very much the same. I will address scoping and prioritization in my next post.
Remember, if you focus on "top of mind" business issues caused by poor data quality, you'll keep your executives engaged! This is the time where you start thinking about what business measurements you can take to measure progress or success.
Oh, and one more thing: every time you mention one of the following words, the "deer in the headlights" factor goes up and you'll risk losing your executives' interest: metadata, data model, business object, ETL, ESB, SOA and data integrity. Talk to your business leaders in their language, not yours!
Leave a Response
Entries(RSS)