US SBA Develops Predictive Analytics to Boost Business Growth

U.S. SBA Develops Predictive Analytics to Boost Business Growth

Much discussion today around data analytics in the public sector understandably focuses on how analytics helps agencies make better-informed decisions that stretch taxpayer dollars and improve mission performance. The story of the U.S. Small Business Administration’s data transformation showcases just how important these benefits can be in accelerating an agency’s impact.

Many agencies are swimming in data, but being able to find relevant and actionable data is the step that needs to be taken, and taken quickly
Sally Block, Vice President of North American Government Solutions, Dun & Bradstreet

Over the last 15 years, SBA has created a data-centric culture to find organizational efficiencies, enhance program outcomes, improve stakeholder relationships, and, ultimately, benefit the nation by expanding its capacity to support small businesses. SBA has substantially improved program performance, realized significant organizational efficiencies, and applied industry best practices in mitigating risk, executing financial decisions, and making those decisions more objective and transparent. In doing this, SBA has generated considerable buy-in and engagement from the small business and lender communities, yielding more loans, more businesses assisted, and more jobs supported and created.


A Data Culture Emerges

SBA’s data transformation began in 2003 when it implemented the Loan and Lender Monitoring System (L/LMS) with the goal of getting a more complete picture of its lenders and loans. L/LMS is a data warehouse that imports a wide array of third-party commercial data about SBA lenders, loan applicants, and recipients on a monthly and quarterly basis. The system immediately expanded the breadth of data available to SBA, enabling it to adopt industry best practices in how it monitored, reviewed, and analyzed its loan portfolios and lenders.

The impact of this new capability was transformational. “By having this data warehouse,” said Sally Block, Vice President of North American Government Solutions at Dun & Bradstreet, “it allowed them to look at loans in multiple ways and select the riskiest of lenders for review, to ask better questions, to look at problems differently, and, subsequently, to look at solutions differently.”

Because their databases had become more relevant to their needs, SBA staff became more attuned to the data and their insights. “They developed a feedback loop,” Block said. “The more data you get, the more inquisitive you become. The more inquisitive you become, the better questions you ask. And the better questions you ask, the better results you get.”

As they collected and analyzed more data, SBA found it could become more predictive and precise in detecting and mitigating risk, said Block. SBA staff gained new insights into the dynamics of their programs and used those insights for better performance, opening new opportunities for the agency and widely affecting agency operations and policy.

The Impact of a Data Culture

SBA’s improved analytical proficiency yielded significant changes at the operational level. For example, the agency vastly expanded the number and variety of risk-based lender reviews it was able to conduct – from 50 to 75 reviews per year to now more than 350. This is because, with analytics, desktop reviews (which are faster and require fewer resources) can be more robust and meaningful. SBA is also better equipped to discern which lenders require more focused, on-site attention from a review team and which ones can be reviewed from a desktop.

Predictive analytics, which employs forecasting models, helps SBA more quickly spot risk indicators exhibited by loan holders or lenders. This enables the agency to more efficiently direct investigative resources and, if necessary, take steps to mitigate risk before it becomes a problem. “I like to say this has put SBA on ‘the front foot’ because it has an early warning system that it has incorporated into all of its loan programs,” Block said. From the perspective of SBA, which operates on a zero-subsidy basis, improved risk mitigation translates into an ability to scale its offerings and, ultimately, put more capital in the hands of small businesses.

Assessing Results

Expanding capacity to meet one’s mission is certainly one key gauge of success for an agency’s data initiative. There are others as well – in the case of SBA, results also arrived in the form of better program performance, accelerated operations, and more efficient resource utilization. Since 2002, the year before SBA implemented L/LMS, the dollar volume of its two largest loan programs – the 7(a) and 504 programs – has more than doubled.

But other benefits shouldn’t be overlooked. SBA’s data transformation also meant it could inject transparency, objectivity, and industry best practices into its vast operations. Adopting industry best practices in 2003 enabled SBA to transparently and objectively demonstrate to its stakeholder base of small businesses, banks, and Certified Development Companies (CDCs) how and why it makes the decisions it does, Block says.

The Journey Continues

The data transformation journey is like most others: First crawl, then walk, then run. SBA has made great strides during the first two stages: putting in place a foundational system to transform around, populating it with relevant data, and applying analytical tools to convert data into actionable intelligence. With objective data at the center of its decision-making, the agency now looks to run. A key focus today is on finding better tools and ways to visualize data to make it more actionable, more quickly, for more people.

SBA’s example offers some helpful takeaways, Block says. First and foremost, understand the data you have. Then understand the data you need in order to make better decisions. “Many agencies are swimming in data, but being able to find relevant and actionable data is the step that needs to be taken, and taken quickly,” Block said. Another lesson: Data by itself is not enough – employ analytics tools that convert data into actionable intelligence. Finally, the creation of a data-centric culture requires persistent, effective leadership. 

As the SBA example demonstrates, the power of data analytics can be truly transformational. There are clear metrics of its success at the policy and operational level. But ultimately these benefits have contributed to a more modern view of the agency and its mission, both internally and externally, and have led to better relationships with stakeholders to create high-functioning public-private partnerships.