Better data management requires more education, communication across enterprise, experts say

In Featured by ThePeopleLeave a Comment

There are certain communities across the federal government that excel at data management, like intelligence or acquisition. Outside of those communities, however, a widening consensus on the value of data has outpaced education on how to use it.

“Sometimes you walk into a room and somebody says, ‘Yes, let’s do data.’ And you literally know that’s the only thing they realize is ‘data.’ And there’s no discussion, there’s no understanding of what that means or the analytic flow that goes from ingesting the data to doing the analytics,” said Lt. Col. Ron Synakowski, director of data capabilities for the Air Force, said during a Dec. 15 GovExec webinar on data management. “And so we’re helping our decision makers and the folks in that community understand those steps so that we can resource that properly.”

In fact, proliferation of data tools across enterprises has in some cases created as many challenges as they’ve solved. For example, Richard Allen, chief data officer at the Environmental Protection Agency, said that before he was CDO, he created a data visualization tool that is now used across the agency. But now, due to lack of a unified data strategy, different offices and programs within the EPA are duplicating data and cleaning it multiple times.

So Allen worked with EPA’s chief IT architect to interview leadership across the enterprise to discover what data challenges they had, what tools they were using, and build all of this into EPA’s new digital strategy.

He said they found a couple of themes during these interviews.

“One is we don’t communicate as well as we could internal. So our people don’t necessarily know where the data that they want are, they don’t even know if that exists necessarily. And even if they do, finding it can be a challenge,” Allen said during the webinar. “And then that there’s a challenge for us in terms of providing the proper guidance. So if there’s a lack of overall guidance from us, people create their own governance strategies and create their own management. And we don’t want to throw out the baby with the bathwater. We’re working with them to figure out what their best strategies are, and how to integrate them a little more holistically across our agency.”

That’s why Allen said he’s been focused on working with the data governance body that CDOs were required to stand up. He said another council of executives wouldn’t help him get the support and labor he needed to do the work that’s necessary. And since he doesn’t have a staff, he’s used the data governance council to pull in the people that he needs from other offices across the enterprise.

“My focus has largely been around building a coalition of people who are our data field agents, if you will,” Allen said. “They are the heroes, and I need their buy in and their help to really help us get cradle to grave data management and governance strategy and procedures.”

The council is also helping him to understand how people across the agency are generating and using data. They’re currently in the process of understanding the agency’s many information collection requests, where that data flows to, when it’s being duplicated, and why. Because sometimes, he said, statutes require the duplication of data.

And then there’s the question of generating better quality data.

“Quality is a multifaceted thing,” Allen said. “One person’s need for quality is different than another, depending on the analysis, timeliness, continuity, all these things matter. But they don’t matter all the time, they don’t matter equally. So the challenge for me there is to first figure out how we’re collecting these data, [and] what the quality elements are of those data.”

So for example, he said he might have some raw version of what s statute considers “authoritative data.” But there’s usually another version of the data that’s more useful to people, depending on their needs.

“So when somebody says, ‘What is your authoritative data?’ Or ‘what is your best quality data?’ Well, ‘what do you want to do with it?’ is the real question you have to ask back to them,” Allen said.

Leave a Comment