In addition, data plays a critical role in building resiliency in the supply chain, he says. Basic data essential to risk management includes knowing who your suppliers are - not just tier one suppliers, but also tier two and three - and where their production facilities are located, says Rotella. "If your supplier's plants are in a hurricane zone or a zone that is prone to flooding, then you can factor in that risk and build resiliency into your network, whether through redundancy or contingencies," he says.
Verisk Analytics has been doing catastrophe models since 1987, Rotella says. "We model earthquakes, typhoons, hurricanes, tornadoes, floods, fires, pandemics and terrorism in about 100 countries. Using those probabilistic models we can help companies understand potential perils that can impact their supply chains."
Corporations' awareness of the need to do this type of modeling has been heightened by recent disasters such as floods in Thailand, the tsunami in Japan and Hurricane Sandy in the U.S., he says. "It is an emerging area but interest is rising. The challenge is in putting a dollar value on the risk."
Verisk also has systems that track other risk elements, such as those associated with global sourcing and extended supply chains. "We do what we call risk adjusted optimization," Rotella says. "It is not good enough to have an offshore component of the supply chain that optimizes costs. You also have to account for the risk elements, and we introduce those as part of the optimization."
These risks include geopolitical events, says Rotella, noting that Verisk focuses on predictive modeling. "When you get a news feed about an event it is almost too late to react, so we look for data sources that can be correlated to other elements to predict potential unrest or problems before they happen." These correlations are not necessarily obvious. "The key is to gather as much data as possible and let data scientists see where there may be correlations."
There are tremendous amounts of data out there and plenty of methods to glean insights from them, says Rotella. "But it's critical to put the data in context; that's when you get really powerful results."
