In today’s information-driven economy, data is extremely valuable to just about any industry and profession, and risk management is no exception. Businesses that are able to properly harness data can apply it to improve their operations and make more efficient use of resources.
“With the rapid growth in computational power and more readily available artificial intelligence tools, there is an increased desire across many businesses to develop tools that can help improve decision making,” said Pentti Tofte (pictured above), staff senior vice president, data analytics at FM Global. “At the core of all this is the need to have high-quality data, and lots of it. Without good data, it will be very challenging to get your analytics and/or AI efforts off the ground.”
This rapid increase in value has led to data being compared to gold. Tofte highlighted the huge contributions of advancements in data-processing capability in working with data. This capability has grown by leaps and bounds compared to 10 or 20 years ago.
“Both availability of capacity and sheer computational power have increased dramatically in the last decade,” Tofte told Corporate Risk and Insurance. “The bar for entry to access this power has also been significantly lowered by virtue of the wide availability of cloud computing, which reduces the need to have in-house expertise and hardware needed to process large amounts of data. The availability of computing power and abundance of new data sources are the two main drivers of the rise of the AI applications that are available today, such as automatic recognition of damage in imagery or automated small-claims processing.”
Due to the huge value now being placed on data, there are concerns that a “gold rush” regarding data could emerge, leading to chaos and a free-for-all attitude.
“The free-for-all attitude may manifest itself in a slightly different form,” Tofte said. “It’s easy to run into problems if the talent pool you’re hiring from does not have the experience and education to spot problems when they arise. It’s becoming easier to build and deploy AI models because of off-the-shelf tools that allow you to do that, but if you don’t understand, for instance, how an AI model can become biased, you have a problem, and even worse – you may not know you have a problem. That’s why at FM Global, our data scientists typically have a PhD in statistics, math or engineering. That’s a stringent requirement we take seriously, and it does narrow our talent pool. But we want to make sure the teams have the knowledge base to apply the proper rigor when working with this technology.”
Tofte said that the most straightforward approach to harnessing data and beefing up their climate risk management capabilities is to partner with a company with expertise in the area.
“When companies partner with FM Global for their insurance program, they receive worldwide engineering and loss prevention servicing which focus on mitigating not only fire and equipment risk, but also climate-related hazards,” Tofte said. “Annually, FM Global’s 1,300-plus field engineers collect data from their evaluations of more than 60,000 locations worldwide that give valuable insight into factors that can affect policyholders’ resilience to property-related risks like the changing climate.”
Tofte expects that the importance of data and analytics will only continue to grow in the near future. This makes it important for organizations to have the right mix of people working for them or with them.
“You should have data engineers to manage the data, data scientists to develop AI models and business analysts who understand the business.” Tofte said. “These roles have to work together to take a business problem and translate it into an analytics solution. Not every organization will be able to capitalize on the full breadth of benefits that high-quality data and analytics efforts afford you – but for those that are prepared to make the commitment to strive to be a data-driven organization and foster a culture that embraces the importance of data, they will see returns on their data practices that far outweigh their investments.”
