5 Tips for Data Management from Federal Leaders


A few federal data leaders shared their thoughts at a recent panel on financial regulation.

Data has become a hot tech topic in the past couple of years, but federal leaders still struggle to communicate the importance of data management. 

At a recent conference in Washington on financial regulation, a handful of federal officials offered some thoughts on the federal data landscape. 

1. Agencies Should, but Often Don't, Share Data Standards

“Standards that are universally adopted also help reduce the burden, both for those who have to file information and for those who have to receive information,” said Linda Powell, the Consumer Financial Protection Bureau's chief data officer, during a panel discussion at the Data Transparency Coalition's financial regulations summit. “It makes it easier for information to flow through the financial ecosystem.”

But sharing financial data between agencies for cross-agency analysis is often full of headaches, said Darrell Ashton, the Federal Reserve Board of Governors' section manager for data management. That's because agencies haven’t yet implemented shared standards.

What results, Powell said, is a wasteful cycle in which agencies bring in data, transform it to fit their own purposes, analyze it, and send it out again for the next agency to repeat.

“You end up with a lot of trains unloading and reloading,” she said.  

Interagency data sharing is a challenge, but standards are just as important for internal data processing, Ashton added. “The more standards we have internally, the more efficient we can be in that process, and reduce the burden to hopefully provide better, more high-quality data.”

2. Data Standards Aren’t Enough

Another challenge, Powell said, is making sure groups filling out financial forms, or later processing the data, are using the same definitions to contextualize data. Terms such as “capital” and “net-worth” can connote different concepts, depending on the interpreter.

The challenge, she said, is “ensuring that we’re all working off a common dictionary."

3. Federal Agencies More Willing to Talk about Data Today

When Donna Roy, the head of the Information Sharing Environment Office at the Department of Homeland Security, joined the agency eight years ago, her chief data architect position was one no one wanted, Roy said.  

“It took me three to four months to get in with the CIO,” she said. “I’ve seen the evolution of conversations in data . . . Now, when the phone rings and my name’s on the other end, leadership in the department picks it up because they understand that [we’ll talk] about the use of data, the value of data.”

4. Federal Government Still Needs to Catch up to the Private Sector in Data Management

“Corporations that are making money, they’re doing it really well . . . and they’re governing their data,” Roy said. “We’re doing it in pockets and we’re governing it in pockets.”

Within the federal government, “as soon as you say 'data management,' people zone right back out,” Roy said. “If you focus on the data analytics or the business problems you’re trying to solve, they will talk to you . . . but we still have a long way to go on institutionalizing" data management within government. 

This requires leadership to communicate their data needs in a nontechnical way, Powell said. Once, she said, she checked out a stack of marketing textbooks to help her better “market” the importance of data management to her agency.

Eventually, she came up with the analogy of a soup can with its label ripped off.

“The soup is the data, the can is the database and the label is the metadata,” Powell told colleagues. “If you don’t have metadata for your data warehouse . . . picture your pantry with a whole bunch of cans in it, but you’ve ripped all the labels off your cans . . . I’m not going to starve to death, but I won’t necessarily get what I want.”

5. The Best Data Hires are Creative

When hiring staff, Roy said she looks for creative, critical thinkers. "I teach everything else," she said. 

Her team at DHS includes visual artists, performing artists and people not trained in IT.  

"They have a way of thinking that translates into the ability to leverage data for the business problems we’re facing today.”

(Image via HerrBullermann/ Shutterstock.com)