In a two-week period during the COVID-19 coronavirus pandemic, the Small Business Administration (SBA) processed more total loans than it had in 14 years, according to Melvin Brown II, director of the Enterprise Business Management Office within the agency’s CIO Office. At the May 13 FCW Workshop on Data and Analytics, Brown shared the data management tips that helped the agency succeed.
With stakeholders in 15 offices, Brown said that SBA is working on using data to develop a 360-degree view of the customer and “create a trail, or process, that allows us to be customer centric, that we look at the customer first, that we’re focused and informed on the customer.” Ultimately, the goal is to develop an enterprise customer relationship management system driven by data to empower employees around customer touch points, he added.
The system will allow SBA to have goals, metrics, and data integration around customer relationships, Brown continued. He explained steps that agencies must go through to fully take advantage of the data available.
“The first thing that you want to do is get organized,” Brown said. “You want to get your data organized in such a way to where you can make decisions.” He explained that data collection alone is not enough – it must be organized properly to build the right foundation.
Next, Brown said that data visualization can help the office look for connections, anomalies, and trends to help further understand the data. From there, he continued, analysts can begin to examine the relationships within the data.
“What I mean by relationships with the data is: where are their commonalities, where are other common themes, where are their touch points,” Brown explained. For example, he said that with the data about who is applying for small business loans during the COVID-19 pandemic, the agency can see the geographic makeup of where loans went to help inform policy decisions.
Brown said it also was important to look for outliers to understand what they mean within the context of the rest of data. He advised others to make sure that their data is coming from valid sources, keep the data protected, not to try to manipulate the data to tell a certain story, and to collect metadata.
