Speed Modernization with Predictive Analysis

Government agencies are data-rich organizations, and it’s time to leverage that data via predictive analytics to drive innovation and modernization. Agencies will never have unlimited resources, so it’s important to adopt force-multiplying technologies and methodologies to make significant progress in mission performance. Modern agencies are agile and quick to adapt to change; and, as we see in our day-to-day work in government, the most advanced agencies are able to anticipate change.

Yet many agencies are still unsure where to begin and how to use predictive analytics to better achieve their mission goals.

Start With the Problem

Agencies should first begin with specific mission objectives in mind and determine what kind of data and insights could accelerate mission progress. Once that is defined, they can plot the steps needed to reach those objectives. For example, law enforcement agencies need to stay ahead of criminals to identify suspicious behavior and illicit patterns before illegal activity takes place. In the case of fraud prevention, analysts can begin by looking at data that reveals trends in previously observed fraudulent activity and connections between businesses and individuals.

Many government professionals are familiar with using data to answer the questions “What happened?” (the descriptive phase) and “Why did it happen?” (the diagnostic phase); far fewer have the tools to answer the questions “What will happen?” (the predictive phase) and “What should I do?” (the prescriptive phase).

In terms of this colloquial analytics maturity model, most federal agencies are still operating in the early “descriptive and diagnostic” stages. They’re challenged to figure out how to aggregate all that data and convert it into actionable intelligence. The Small Business Administration (SBA) is a great example of an agency that’s ahead of the curve in their adoption of modern tools, specifically predictive analytics, to further their mission. For example, SBA was the first agency to use credit scoring in a meaningful way, at a time when it was rare for government to use predictive modeling. The efficacy of SBA’s predictive models in managing loan portfolio risk has certainly played a role in the agency’s status as a zero-subsidy organization that does not rely on appropriated funds to run their lender programs.

Reformulate the Analytics Maturity Model to Build Toward the Solution

Every agency is at a different stage in the development of its data and analytics capabilities, in terms of how they’re leveraging tools to move up the curve.

At Dun & Bradstreet, we’ve updated our way of thinking in this space. Instead of the descriptive-to-prescriptive paradigm, we’ve developed an analytics maturity model specifically for government agencies:

This analytics maturity model begins with smart data, which is analytics-infused data that can be leveraged for actionable intelligence; it has context. For example, Dun & Bradstreet’s Financial Stress Score comprises data within our database, such as payment, public filing, demographic, and financial information, to help predict the potential insolvency of a business. The next phase is predictive analysis, or rule-based models that can make a major impact on mission outcomes. Predictive analysis takes us to a stage where data helps identify the likelihood of future outcomes based on past behavior. For example, regulatory agencies are developing scorecards and watch lists to detect fraud signals and prioritize time-intensive investigations, which often include physical inspections, to better focus on high-risk endeavors.

Finally, mission-based analytics is where smart data and predictive analytics are truly ingrained into every single function of a government agency. Its goal is to provide anticipatory analytics and enable agency leaders to ask questions that haven’t been asked before – not only to better protect the nation from threats but also to spur innovation and modernization.

The Return on Investment? A Modern, Data-Driven Agency

It’s possible for every agency today to reach the mission-based analytics stage. The answer isn’t more processes or people – agencies must find ways to do more with less. In some cases, a “boots on the ground” approach is necessary and human intelligence is irreplaceable. However, field agents will be much more effective and efficient in meeting mission objectives when they have analytic information to inform and guide their decisions.

Mission-based analytics begins by choosing a very specific mission objective to tackle, first by collecting smart data and then by applying analytics to that specific objective. Leveraging such predictive analytics can lead to better use of limited resources by focusing efforts on mission-focused work rather than mundane tasks, becoming what is known as an “agile enterprise,” once the domain of the private sector.

Consider the Food and Drug Administration (FDA) and Customs and Border Protection (CBP). Their agents inspect food production facilities and shipping containers for suspicious cargo, respectively. CBP also has a tariff and revenue objectives to make sure that appropriate tariffs are collected on imports. But in order for those agents to be most effective with inspections, they need to rely on predictive analytics to “reduce the haystack,” which guides their searches to focus on only the riskiest cargo containers.

The FDA regulates a sizeable percentage of imported consumer products – food, pharmaceuticals, medical devices, cosmetics, and tobacco. They can’t physically inspect every single import that comes into the country. So they use analytics to determine proclivity of suspicious cargo based on the container’s ownership, country of origin, and other factors. That analytics-infused data, combined with surveillance from human intelligence, helps to find and root out any counterfeit, unsafe, or inauthentic items in the name of national security and public safety. Modern tools for an agile enterprise.

Time is of the essence when it comes to implementing modern tools to create an agile agency. As governments continue to look at smart data and predictive analytics to expand and scale, analytics should play a large part in supporting agency outcomes for a better future.