Commercial insurance underwriting is being transformed by the greater availability of detailed data and automation.
Information gathering previously done manually by underwriters, brokers and policyholders is increasingly obtained in a fraction of the time once needed through technology and is prepopulated into coverage applications.
Using internal data and working with third-party data providers, insurers are using artificial intelligence and other technology to make underwriting faster and more accurate for small and medium-sized commercial risks.
Such efforts are “more pointed at small business today, but I do think the same sort of technique also will benefit the middle (market),” said Tony Fenton, Nationwide Mutual Insurance Co.’s vice president of commercial lines digitization and innovation in Des Moines, Iowa.
Automated underwriting is “taking what I did 23 years ago and finding a better way to arrive at a decision,” he said.
Automation can reduce much of the “heavy lifting” and mundane, repetitive work of gathering information, said Dogan Kaleli, New York-based co-chair, global facilities/programs practice group, and head of programs business, North America, for Allianz Global Corporate and Specialty, a unit of Allianz SE. “That manual process got very frustrating and time consuming.”
FM Global uses robotic process automation to “try to remove some of the soul crushing work involved in taking data fields from one form and entering them into another manually,” said Michael Lebovitz, senior vice president of innovation at the Johnston, Rhode Island-based insurer.
Underwriters are freed up from having to do basic data processing work, which gives them the opportunity to think more about risk, said Sean Ringsted, Bermuda-based chief digital officer at Chubb Ltd. “What’s different today is the extent of the capabilities that are available and the speed at which you can do things,” Mr. Ringsted said.
Data, such as location information, building values and other municipal information, that is needed for insurance submissions can now be collected automatically from a range of sources, Mr. Kaleli said.
Karnataka, India-based Bridgei2i Analytics Solutions Pvt. Ltd. works with insurers to help bring automation to underwriting and other processes, said Aakash Desai, Boston-based head of banking, financial services, insurance consulting and business development at the technology firm.
“Using data crawling techniques, we are scraping the web for public information or paid subscription websites which collect third-party data,” Mr. Desai said. The system “uses data much faster than humans,” and helps to create efficiencies in the data entry part of underwriting, he said.
Another advanced data analysis method used by insurers is “computer vision,” said Mr. Desai, who described it as “a set of techniques used to review information in the submission; used to capture words, images, and any information within documents.”
Other sources said similar techniques can be used to analyze and gather information from images such as photographs and video.
Tapping data, both publicly available and that from paid subscription, is becoming more widely used in the insurance sector.
“We’re using technology like pre-filled data to help reduce the back and forth with the policyholder,” said Denver, Colorado-based Mike Ferber, CEO of managing general agent Dovetail, a unit of Victor Insurance Managers Inc., the underwriting management arm of Marsh & McLennan Cos. Inc.
“To the extent there is data out there, we want to access it. The data is getting better every day but it’s not nearly as ubiquitous as the personal lines data,” Mr. Ferber said.
Dovetail uses a bespoke connection with each insurer, which usually have slightly different systems, “and therein lies a lot of the work.”
“We are looking at digitizing and automating a lot of our products, underwriting processes, building self-service capabilities,” said Sumeet Bhatia, Chicago-based head of innovation for Zurich North America.
Zurich is also “looking into how we can continue to build (application programming interfaces) to integrate with a digital focus” because many of its distributors also are building digital capabilities, Mr. Bhatia said.
Data from partner entities such as third-party administrators can help inform decisions on risk, said Doug Alexander, enterprise architect, Axa XL, a unit of Axa SA, in Birmingham, England. “We can use that data to improve risk engineering and underwriting.”
“The biggest impact of technology has come from the data and the ability to leverage it,” said Nate Kraynak, digital product manager, Chubb Small Business, based in Malvern, Pennsylvania. “This aspect of technology has really streamlined the process and made life easier on all parts of the value chain.”
“In the old days, it was a more painstaking process for that business owner to come up with that specific information and many times it was a lot more unreliable as well,” said Mr. Kraynak. “Now, we get that information from third-party sources and in many cases it’s more accurate.”
Automation in underwriting works most effectively in the small business sector, for example, companies with annual revenues of $5 million or less and/or those with 10 or fewer employees.
One firm supplying data to insurers and others is WeatherCheck, which provides historical weather information for a given address or structure, said Demetrius Gray, founder and CEO of the San Francisco-based company. Users can type in an address and “get a severe weather history associated with an asset.”
Founded three years ago, WeatherCheck started its service with hail weather patterns and has since added a wind model and supplies its data to surplus lines insurers and others, Mr. Gray said. Data sources include public sources such as The National Oceanic and Atmospheric Administration, part of the United States Department of Commerce, as well as paid subscription services.
