Clean-up Is Thankless Job; IT Should Take Lead in Data Governance

Share it on Twitter  
Share it on Facebook  
Share it on Linked in  

In many if not most instances, you (or a family member or friend) will become aware of a health issue before your doctor does. However, early detection won't help you if you aren't motivated enough to see a medical professional about it.


The same principle applies with data quality. No one in a company should ever react with a shrug when they see data discrepancies or other errors. Yet all too often that's exactly what happens, because folks assume "someone else" will fix it or believe it simply isn't that big of a deal.


I wrote about the idea of bringing more users into the data governance process a few months ago, noting that it's more difficult to deal with problem data after it's been disseminated multiple times by multiple users in multiple systems. I suggested a hierarchical system proposed by Gartner analyst Ted Friedman in which all users have a stake in data quality, but only data stewards make process changes or commit resources to address issues and a corporate data sponsor (probably someone in IT) serves as the ultimate authority.


A post from Charles Blyth on his MDM from the Front Line blog makes a fine follow-up, I think. In it, he defines five user groups that should be involved in data governance and examines their roles in more detail. They are:

  • Authors: These folks create the data and are thus closest to it, yet they generally don't feel any ownership and are concerned only with the data as long as it remains in their immediate scope of responsibility. (In other words, not long at all.) Blyth says it's "imperative" to get those folks more involved with data governance.
  • Consumers: They are the heaviest users of data, so should play a key role in catching data quality issues that went undetected or unaddressed as data was entered. This way, problems don't move farther up the data chain.
  • Enrichers: They augment or change data. Like authors, they need to become aware of the broader ramifications of letting poor quality data leave their control.
  • Publishers: These people must be perfectionists because it's their responsibility to provide content to consumers and other business users. They are often the folks that bear the brunt of user complaints.
  • Custodians: This is where IT tends to enter the picture, with database administrators, data architects, data stewards and data warehouse developers. Custodians have traditionally been responsible for cleansing data and establishing data standards. But their role needs to broaden, writes Blyth, extending to "defining the data governance strategy with the business, managing the deployment and support of that strategy, developing the business data model and supporting the IT architecture within the data governance domain." So, custodians shouldn't just clean up other people's messes. We all know that's a pretty thankless job. Instead, they need to take a more active role in getting everyone invested in keeping corporate data repositories as clean as possible.


How to get everyone more invested in the data governance process? Blyth suggests establishing key performance indicators centering on data quality, providing incentives for meeting KPIs, and communicating the importance of data quality as often as possible, connecting data quality with the improved ability to meet business goals.