6. New approaches to underwriting risk and predicting loss This theme represented 3% of applications to Startupbootcamp, the We have already begun to see more start-ups that are captur- smallest category. Some 54% of corporates considered this theme ing and analysing data (see theme five) that insurers can leverage impactful, ranking it as the third most impactful of our six themes. in order to refine or re-engineer their underwriting processes. Fur- Traditionally, insurance has operated on a protection model ther growth will come if start-ups are able to access deep industry where it supports policyholders following a loss. However, we are knowledge and combine this with improved data capture and anal- now seeing a move towards more advanced, preventative models ysis and machine learning. A partner-led model where insurers and that enable insurers to mitigate such losses. Real-time data capture new entrants work together will be the key to success. “Alone, start- from sensors and advanced analytics enable insurers to predict an ups still lack a number of essential competencies that incumbents impending liability and underwrite in more precise and sophisticat- have built up over the years,” says Martin Pluschke of ERGO, while ed ways. Rod Willmott of LV= adds: “The start-ups lack historical data without The small number of Startupbootcamp applications in this area partnerships and in some cases there is no history for what they reflects the very technical nature of underwriting, the extensive are trying to do.” data required, and the need for actuarial expertise. For these rea- Discovery provides a good example of the potential for partner- sons, insurance incumbents have a competitive advantage. That is ships. It is a health-focused insurer that has partnered with Human not to rule out the start-ups. “Many of the incumbents’ data models Longevity, which is able to provide whole genome and cancer ge- are somewhat outdated,” says Dan Smith of MMI Holdings. “What nome sequencing to clients. Customers benefit from personalised start-ups introduce is alternative data types, sources and time se- health and medical treatment strategies and will be able to better ries that can enrich the existing incumbent underwriting process as manage or avoid serious health problems. opposed to competing with them.” 25