Typical "worker bees" are incredibly valuable to any organization. They show up on time, speak when spoken to, get their jobs done and go home at 5 p.m. Just don't ask them to join your master data governance team.
Instead, a master data governance team should consist of an organization's most capable, energetic and charismatic leaders, because they will be called upon to make and enforce unpopular decisions, according to speakers at last week's Gartner MDM Summit in Los Angeles.
"You’ve got to get the right dynamic people," said Joe Young, an information delivery manager at computer printer manufacturer Lexmark International Group Inc. in Lexington, Ky. "When you go from little or informal data governance to formal master data governance across an organization, you've got to have people that [can] stand up and say, 'That's not the way we do it anymore.'"
The phrase "master data governance" refers to the policies, processes, controls and audit functions used to manage and secure an organization's critical master data. "Master data" refers to an organization's common business data entities -- such as customer, product and supplier -- and their associated values and attributes.
In a 2010 blog post, Gartner research vice president Andrew White explained that master data governance provides a framework where processes are maintained that assure a single and consistent view of master data whenever business workers need it.
"Master data governance does not care for any one specific need; it has to care about needs for all uses – analytical, operational, transactional, whatever," White explained. "However, master data governance does not [assure] the right business decision is taken [or that] the actual desired business outcome is achieved. [Master data governance] is an enabler, not an end to itself."
Selecting a master data governance team
Many large-scale master data governance initiatives are headed up by councils that are charged with overseeing everything from defining master data to implementing programs to establishing policies and procedures. For example, a master data governance council might decide to implement a standard corporate hierarchy for master data management in an effort to increase consistency across business units.
But such decisions are often unpopular with business units accustomed to a higher level of autonomy, Young warned. That's why it's important to have master data governance council members who are skilled in the art of diplomacy.
The data management council will be charged with examining all important data and processes related to key entities such as customer and supplier. But they should also come to closely understand any interdependencies that cross multiple domains, added Jodi Maciejewski, the director of data governance at ConAgra Foods Inc. in Omaha, Ne.
"It's not just looking at the data and processes for each area," she said. "It's looking at how they relate to each and what are dependencies of those processes."
When looking for the right "master data governance council" members and project leaders, it's a good idea to start at the top of the organization, Young said. At Lexmark, which instituted a master data governance program three years ago, that meant getting the chief financial officer (CFO) on board right away.
"We started with [the CFO and] he got engagement from the line of business vice presidents," he said. "That's when we formed a council."
The Lexmark master data governance council then met regularly for six months to identify and "scope out" master data problems and begin generating mutually agreeable recommendations as to what needed to change.
"We hadn't bought any tools. We hadn't written a line of code. We hadn't analyzed one record," Young said. "Before we started anything, we got that council set up."
Council members should be flexible and prepared to adapt as master data governance policies change along with business objectives, Maciejewski said.
"You may have a plan that you think is going to work, but as you get in and start executing the plan you find out that you're not structured the way you should be or you're not looking at the right processes [in the] right way," she said. "One of the things that we've learned is that you need to be flexible."
Get business units to speak the same master data governance language
A good master data governance council should also set out to make sure that every business unit is speaking the "same language" when it comes to master data, said Jolene Jonas, a data architect with Santa Clara, Calif.-based computer processor giant Intel Corp.
Prior to implementing Intel's internal master data governance program, the company gathered key stakeholders in a room to hammer out the precise meaning of words like 'customer' and 'supplier.'
"It's so critical in governance that we're all always talking the same language and it's consistent across the enterprise," Jonas said. "If I have the key stakeholders together buying off on that, that's a [great] first step."