Data Governance: Start Where You Are

Loraine Lawson

Do you know who cares about data governance? Not budget-focused executives, that's for sure. Not the business users who, really, should own the data and stand the most to gain by getting it right. And not even middle managers care.

 

If these people, who really are the ones who wanted the data in the first place, don't care, who does? That's right: IT, particularly the database managers, data architects and others in IT who have to clean up the tech messes caused by bad data, according to a recent TechTarget article.

 

What's wrong with this picture?

 

I could go on and on about how it takes executive sponsorship to make data governance work. I could list the numerous other, expensive undertakings that rely on good data to achieve ROI, including on integration-heavy projects such as customer data integration or product information management, data quality and master data management (where data governance does, at least, enjoy some success), and business intelligence. I could explain the obvious, which is that if the business is the one using the data to drive decisions, shouldn't it at the very least want to ensure that data is as reliable and secure as possible?

 

But it probably wouldn't matter. Heck, just by mentioning data governance, it seems I'm at risk of losing all but the choir, and what's the point of preaching to them? According to the SearchDataManagement.com survey cited in the article, a third of organizations aren't addressing data governance at all, and those that do, tend to do so in an informal way, at the departmental level.


 

So the question for IT becomes: How do you sell data governance to the rest of the organization?

 

And before you post to tell me IT shouldn't have to "sell" anything, let's get real here-it's certainly not catching fire on its own, is it? Right now, data governance ranks right up there with a root canal. And you can't possibly believe business shouldn't be involved in data governance. So, if you don't like the idea of "selling" the business case on data governance, call it lobbying, call it a public service announcement, call it whatever you want, but can we all just agree here that this is something the business should care about and own on some level?

 

Ideally, data governance should be a top-down initiative with strong executive sponsorship. Forrester's Rob Karel says executive buy-in is the big key to success with data governance. He recommends focusing on the business process to increase momentum.

 

But sometimes, you've got to start where you are, and realistically, data governance tends to be a bottom-up effort. IT should be able to work with that-after all, you've had practice with other initiatives that started in IT and then spread out to the business, including project management and computers in general.

 

Suggested C. Lwanga Yonke, an information quality professional and advisor to the International Association for Information and Data Quality, in the TechTarget article:

I think that oftentimes, practitioners want to start too big. I think it's important to define a specific problem one is trying to solve [and start there].

Starting small might not be the wholesale attitude adjustment the business needs right now, but at least it's a step in the right direction.



Add Comment      Leave a comment on this blog post
Sep 16, 2010 5:40 AM Ed Gillespie Ed Gillespie  says:

As you know Loraine, the hardest step is taking the first one.

In a recent interview, Navin Sharma remarked how data profiling is the obvious first step-so you could collect aggregate statistics about the quality of the underlying source data. An in-depth profile of your data will provide counts on what percentage of the fields are populated, in addition to providing a true understanding of the quality of the data by examining the data values.

A good profile will ask questions such as-How many unique keys are not unique? Are there symbols or commands where characters should be? Are the numbers in an appropriate format? Are fields, such as Social Security number, populated with all ones or all X's? By comparing the universe of values within a database, you can identify outliers, anomalies and other questionable data.

Seems like that arming yourself with facts is a good place to start.

Reply
Sep 24, 2010 3:19 AM Max Gano Max Gano  says:

Hey Loraine, your points are very well taken. Too often "Data Governance" is presented as a monolithic concept which, to executive management, quickly translates into monolithic costs with poorly defined business value.

Starting small and winning big is a proven strategy for those of us who are providing guidance to companies launching efforts to govern data.

And many of us are beginning to speak in terms of "governing data" as a new set of activities added to existing business processes.

Governing data to improve business performance or conformance just makes sense. And that can easily begin locally then grow globally as value is demonstrated. Business people are smart. They want to do things that make businesses run better. And they tend toward incremental adjustment rather than broad sweeping change. Processes for governing data is best introduced with that in mind.

Great post, hope to see more on the topic.

Reply

Post a comment

 

 

 

 


(Maximum characters: 1200). You have 1200 characters left.

 

null
null

 

Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.