'Digital labor' poised to refine underwriting
Insurers are consuming an ever-increasing amount of data from internal and external sources in their quests to more accurately price risks. But that's putting a lot of pressure on underwriters.
The solution is likely to come in the form of "digital labor" to help these professionals get the correct insights out of data. That's according to a recent report from KPMG, "Enabling the Future of Underwriting: A Digital Road Map."
KPMG says technologies like robotic process automation, artificial intelligence, machine learning and cognitive computing won't fully replace human underwriters, but will help make them more efficient by being able to quickly analyze lots of unstructured data and provide scores that allow underwriters to be more assured of a risk profile.
"If you really evaluated what an underwriter does, a lot of it is data collection, cleanup and assessment," says Mike Adler, KPMG insurance principal and a co-author of the report. "Much of that can be taken care of by digital labor and artificial intelligence. So the underwriter of the future may need a different skill set."
Lines of business where products are less commidified, like life insurance or commercial, are likely to be where many of the early advances in using these technologies occur, according to Adler. That's because the more complex the risk, the more advanced the computing technology has to be in order to automate the analytics around it. Personal auto and home products don't require the same nuance.
"We see this happning primarily in life, as well as commercial insurance, to leverage cognitive and AI bringing primarily unstructured content into the underwriting process to better price and predict risk," he explains.
While these technologies are in their infancy, senior underwriters will play a crucial role in training the technology itself to operate within the insurance framework, Adler says. But once the automation is able to process the data at a high level, any underwriter can take advantage.
"The first thing you typically do is train the content so it understands the insurance industry and guidelines around underwriting," he explains. "If you have, for example, your most senior underwriter doing that, one of the things you're able to do is provide that knowledge to a junior underwriter, making that level of individual much more effective."
Adler notes that there are several tests in-market, mostly focused around helping people get comfortable with using the technology on a small scale.
"The trend is toward thinking about underwriting end to end, but starting in a line of business or in a geography as a starting point to design the solution architecture and operating model," he says.