Craig Sanders, president of Enfathom, the business intelligence and analytics solutions division of global consulting company North Highland, has been helping its clients use analytics to increase revenues, cut costs and strengthen decision-making for two decades. Today, he sees new ways to leverage analytics to improve risk management. He recently took time to respond to several questions via e-mail chat.


What are some of the most effective ways you currently see finance and risk managers leveraging Big Data to strengthen their risk management capabilities?

Craig Sanders:Among many Big Data risk management trends, I have seen two new movements that are appearing across a variety of companies. The first involves moving from statistical sampling for managing transaction risk to complete population analysis. With stronger analytic tools and the ability to calculate across large populations, companies are doing great work analyzing risk with the ability to implement solutions quickly against those populations.

Second, new visualization tools put understandable information in the hands of decision-makers quickly and effectively. With strong data discovery reporting tools and visualization capabilities like link analytics, users can point to the presence of risk behaviors without having to go through coding exercises. The analysis becomes a point-and-click effort.


Who are some of the key players who typically collaborate on this analytics work?

Sanders:The three key teams that need to collaborate to make these projects successful include the CFO’s team, the CIO and business leadership. The CFO team brings to the table the insight regarding risk exposures and how the company should prioritize risk efforts. The CIO team brings the strength of understanding data structures and tools that can be deployed to support these efforts. Finally, business leadership, including risk managers, bring insight into the business processes being analyzed and are clearly critical to implementing improvements.


Are there any common pitfalls finance and risk managers should recognize and avoid when integrating more powerful analytics capabilities into their risk programs?

Sanders:One of the biggest pitfalls in these efforts is trying to design and then build the perfect engine in the first iteration. By their very nature, risk analytics involve hypothesis development, data gathering, statistical modeling, testing and tuning analyses, piloting solutions and implementing new routines. Too often, companies try to design a data model that will anticipate every question. That’s almost an impossible task given the breadth of insights possible and, with new discovery-based analytics tools, it’s really not necessary.

A better approach is to establish a hypothesis first, perform analytics in a dynamic environment and then, once the analytics start to show value, consider how to migrate the system to production.


What are some key considerations and practices that enable risk-related analytics endeavors to thrive?

Sanders:The first thing I would consider is strong business sponsorship and a valuable target for analytics. With a sponsor who is driven to improve an organization using analytics, success is more likely to be achieved. Second, think out of the box. There are visualization tools that enable business users to see data in different ways and to understand relationships at a deeper level. This is critical to success.