That’s what turns data analytics platforms into decision-making engines, a pair of government technology leaders said.
Big data and analytics are popular buzzwords in government today, as agencies look to put the huge amounts of data they collect to good use. But, as with so many things in technology, progressing from collecting and organizing data to making better decisions with it tends to be more of a people problem than a technological one.
Agency leaders who want to empower their workforce to use data for decision making should ensure their employees are able to use the tools being provided, according to federal technology managers who joined Nextgov for a live discussion in December.
Prior to joining the Energy Department, John Moses, who currently serves as the director of governance and enterprise management services at the Nuclear Regulatory Commission, worked with another agency to procure high-end analytics tools for things like semantic and natural language processing. Today, he’s seeing those same tools being integrated into common applications like Office 365 and being made available to everyone in an organization.
“We used to call it the ‘holy grail’ to find some tool that did semantic processing like the one I described,” he said. “It’s gone from laboratory, high-end, very expensive to commercial and easy to use. In addition to putting your data in the cloud so you can just expand out the processing power … the ubiquity of the ease of using these tools” are changing the way employees do their jobs.
“Producing the report used to be a great accomplishment,” said Guy Cavallo, deputy chief information officer at the Small Business Administration. “What I’m seeing different with the power of the tools today is you produce this report, you have it on the screen in real time, then someone says, ‘But what if we looked at x, y and z?’ And you start drilling down and changing the scenario. So, instead of giving leadership a PowerPoint and pages and pages for every scenario, now you go to them with an open screen and say, ‘Hey, what do you guys want to know,’ and you drill in in real time.”
With this ability, employees with solid expertise in a given area don’t need to be data scientists, as well.
“Because the tools are easier to use today—and I want to see them continue to get easier—if I really have an expertise, if somebody gives me something that I can really drill down and cross-check a hundred different data elements, then I can spend weeks looking for patterns,” Cavallo said.
When looking for the right analytics tool, Cavallo said he abides by the 80-20 rule while giving considerable weight to the customer experience aspect.
“Will it do everything? No. But will it answer 80 percent of our needs and make it as easy as using Excel? Then let’s do it,” he said. “What we’re seeing by shining the visualization light on data that we’ve had for decades, it’s changing our behaviors.”
Cavallo cited the cybersecurity teams in charge of patching known system vulnerabilities. Using data visualization tools, SBA was able to help those teams sort through all the incoming security information to hone in on the most important, timely patches in need of attention. A tool that was too unwieldy to use would have had the opposite effect, he said.