Aviva,has experimented with selectively replacing costly and inconvenient medical exams with predictive modeling of risk based on enhanced data. A study of 60,000 Aviva applicants found that nontraditional data was as effective in identifying potential health risks as blood and urine tests.
In distribution, companies are improving agent recruiting and retention by developing predictive models to identify, select, and retain the best sales performers. In near real time, these companies track metrics such as early customer-lapse rates, disclosure rates (how much information an agent is able to obtain from an applicant), and average policy size. They also design incentives on the basis of these indicators in order to move performance in the right direction.
Many forerunners provide agents with advanced sales-support tools. For instance, MassMutual Financial Group has developed a mobile app that highlights the customers who are most likely to buy a policy with an AI based predictive engine. The system has increased placement rates by 5 percent in 18 months.
SCOR Global Life, for instance, offers a real-time, fully automated underwriting solution called Velogica, which helps life insurers in the U.S. create affordable products for the largely underserved middle market. Decisions are communicated to an agent in less than a minute in 90 percent of cases, and fewer than 5 percent of applications require human intervention. A rules engine collects and automatically processes external data from sources such as the MIB Group, motor vehicle reports, and prescription databases. More than 1.3 million applications have been underwritten to date with the system.
AllLife provides affordable life and disability insurance to policyholders who suffer from manageable diseases, such as HIV and diabetes, and who agree to adhere to a strict medical protocol. Patients get monthly health checks and receive personalised advice on managing their conditions. With the client’s permission, data is pulled directly from medical providers. If a client does not follow the treatment protocol or discontinues treatment, benefits or coverage can be lowered or cancelled after an initial warning. The company assesses its risk every three to six months, rather than just once.
Approximately 80% of the insurance claims filed with Ping An, a large insurance provider in China- are small claims, and the AI system handles about 95% of these. In the first half of 2017, claims losses declined by 7.8%, or roughly $480 million.
Taikang is leveraging a $250 million annual technology budget to create Health Cloud, an integrated insurance and health ecosystem. The company is offering this service to about 1,000 nursing homes, several thousand dental clinics, retirement communities in 14 cities, and around 1,000 hospitals. Taikang’s goal is to connect these thousands of “data islands” to create a single customer relationship management (CRM)
