This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
3. - Solving problems that arise in data governance programs: Read more in this section
- Implementing standard processes alongside data governance tools
- How companies are maximizing big data benefits with data governance
- Improving data quality with MDM and data governance programs
- IT pros discuss how to overcome data governance challenges
- Energize a data governance strategy with the right set of skills
Explore other sections in this guide:
IT representatives from Fannie Mae Inc. and Sallie Mae Inc. took to stages on opposite sides of the United States this month to explain how their organizations launched what have thus far been highly successful data governance programs.
The data governance professionals from the federally backed financial services firms also offered tips on how to overcome some of the major obstacles that arise as organizations seek to implement data governance on an enterprise-wide scale.
Some of the problems that governance teams are likely to encounter include a lack of skills around data quality tools and tactics and a lack of accountability for the state of data throughout the business units of the organization.
What is data governance?
According to the Whatis.com definition, data governance is a set of processes created to ensure the quality, consistency, usability and availability of information throughout an organization’s business units.
But the most dangerous obstacle of all -- the one with the most potential to derail a data governance program -- is steadfast resistance to newly implemented data governance policies and procedures, according to John Mulholland, Fannie Mae’s vice president of enterprise data architecture.
“When you go through these things, no matter what, there is pushback,” said Mulholland, who spoke to a crowded room last Wednesday at the O’Reilly Strata Conference in Santa Clara, Calif. “This is the biggest hill to climb.”
For Fannie Mae, the key to gradually overcoming resistance involves educating employees about the benefits of well-executed data governance and a strong commitment to cultural change.
“In the past, everybody thought data was a commodity,” Mulholland said. “But now you’re seeing that data is not a commodity. It is the blood that is going to flow through all of our business processes.”
Get LOBs together and hammer out the details
It's a good idea to have representatives from all lines of business (LOBs) brought in when launching a new governance program, said Michele Koch, the director of enterprise data management and the data governance office at Sallie Mae.
Speaking to a roomful of attendees earlier this week at the TDWI Solution Summit in Savannah, Ga., Koch said that business representatives may be reluctant to meet at first, but the benefits of doing so will quickly become evident.
In one example, Sallie Mae, which first launched its data governance program in 2006, held a stakeholder meeting to discuss whether the “anticipated graduation date” field of a particular database should be changed. But one of the business stakeholders in the meeting quickly pointed out that there are compliance issues around the anticipated graduation date field and that it could not be changed.
“We’re national across the United States so I really fought to get everybody brought in,” Koch said. “It turned out to be a very good thing for people to communicate with one another."
For more on the data governance program at Sallie Mae
Read what Sallie Mae’s Michele Koch has to say about data governance best practices
Find out how Sallie Mae measures the value of its data governance program
Create a data governance plan and sell it
The task of implementing data governance at the enterprise level -- a level that truly cuts across business units and data silos -- requires a great deal of planning and a touch of salesmanship, according to Fannie Mae’s Mulholland.
The first step is creating a comprehensive plan. But that can be difficult, especially when organizations make the mistake of stubbornly focusing on a single goal.
“The worst thing that you could possibly do is target your end state,” Mulholland said. “What you have to do is measure your progress from your baseline -- where you’re starting from -- because your goal line is always going to move.”
The plan might include the creation of a data governance council with representatives from all LOBs. It may also include the appointment of data stewards and subject matter experts to usher things along.
“We really looked at this from a staging perspective and a plateau perspective,” Mulholland said. “What we did first was define and design what we were doing. We had a plan. We had to sell that plan.”
Consider taking a hierarchical approach
The data governance program at Sallie Mae is broken up into three tiers. At the top is the data governance council, which includes members from each business unit, said Barbara Deemer, the chief data steward at Sallie Mae.
Under the data governance council is the data governance office, which is responsible for administering the program. At Sallie Mae the data governance office consists of Deemer, Koch and a third employee.
The third tier consists of the data governance services team, which is charged with carrying out and monitoring data quality projects.
“Every single member of the formal council is a member of the business community,” Deemer said. “We actually take votes on the council. We meet every two weeks, we have an agenda, we go through the agenda and we publish minutes.”