Best practices for proactive enterprise risk management: “Let the (big) data tell you”
Visual investigations
In addition to using advanced analytics, the most effective predictive investigations are intrinsically visual for heightened risk awareness. Visualization capabilities offer multiple advantages. The ability to visualize risk networks is crucial for performing impact analysis, which is particularly useful for evaluating points of failure for IT network availability and disaster recovery. Graph techniques are especially helpful there because they illustrate links between system components. More importantly, visualizations are necessary for involving domain experts in proactively deterring risk. “When you show somebody a piece of data, you want to create the right perception in their mind of what the underlying data is saying,” Bennett explains. “That’s surprisingly difficult.”
Visualization mechanisms transfigure data’s quantitative significance into images that are “critically important the farther away you get from data scientists to effectively communicate lots of data in a way that should be internalized by the viewer,” Bennett says. Data visualizations may be the sole way for domain experts to understand high-dimensionality data or data at scale for intricate cybersecurity or fraud detection scenarios.
The practical merit of using visualization techniques to proactively investigate risk is readily apparent in the public sector. Without some techniques to visualize what might be the most important nodes in a complex network or the shortest path between two bad guys in a network to shed light on how they’re communicating, it is difficult to communicate the information, if not impossible, Bennett says.
That example is also applicable to the private sector for fraud detection, cybersecurity and even regulatory compliance use cases. The capital value proposition of network visualizations is they illustrate how threats are occurring. Regardless of the use case, this intelligence is almost always necessary for counteracting risk or suitably investigating it. Such visibility is equally important for preventing its reoccurrence.
Real-world applications
The proactive risk management framework Bennett details may well be the most cutting-edge blueprint for assessing and decreasing the presence of organizational risk. Its utility spans everything from private sector entities to some of the more salient public-sector ones on the international stage. “These sorts of things are useful all the way up to a national level law enforcement agency—agencies at the level of the FBI,” Bennett says. The general methodology, however, can help even small and midsize businesses mitigate risk. The four main principles are:
♦ Data management—Organize data according to risk management domains so there’s a uniformity of access despite any existing silos.
♦ Network generation—Produce a network that denotes connections between data elements pertinent to the domain as the first step toward risk detection.
♦ Detection—Utilize advanced analytics to issue alerts for risk concerns based on business rules or the actual data, triggering the investigative process.
♦ Investigation—Illustrate risk networks with visualization technologies for comprehensive investigations and damage control.
The fundamentals of that approach are embraced by a broad assortment of users. “We’ve been supporting law enforcement with precursors of this system for many, many years,” Bennett says. “Everybody from the London Metropolitan Police all the way to the town of Cary [North Carolina], which is 200,000 people and a small municipality. It really runs the gamut including police agencies in the Middle East that have very dynamic security environments, let’s just say, and everything in between.”
With such an established precedent (and more than a few examples of success), it’s almost certain those same principles can deliver similar yields to private sector companies as well.