A recent Willis Towers Watson survey of the US and Canadian property and casualty (P&C) market has revealed that most re/insurers have been implementing advanced analytics at a much slower rate than anticipated.
The broker reported that, in general, companies had either been overly ambitious in their goals for using analytics or had been held back by the realities of day-to-day business and market challenges.
Nearly half of survey respondents identified problems with information bottlenecks as the biggest reason for their lack of progress, in areas where people and systems typically need to interact.
Other top obstacles included handling infrastructure constraints and a lack of sufficient staff to analyse data.
“Certainly, there is no lack of ambition when it comes to P&C insurers wanting to use data and advanced analytics for the betterment of their business,” said Lisa Sukow, director, North America P&C practice, Insurance Consulting and Technology, Willis Towers Watson.
“But by their own admission, most are falling short of those aspirations right now,” she explained. “One way to recondition their resolve is to align strategy and analytics by identifying what’s core to achieving a competitive edge and steer efforts and resources accordingly. They should avoid getting carried away with analytics for analytics’ sake.”
According to the survey, P&C re/insurers had expected to be further along with their use of advanced analytics by now.
For example, 77% of respondent thought they should have been using internal customer data by now, but in reality only 54% are.
WTW also identified similar gaps in expected versus actual use for social media (46% versus 26%) and clickstream (34% versus 14%).
“Many insurers collect customer data as part of an application or claim process, which can be of wider value when suitable analytics are run,” said Nathalie Begin, director, North America P&C practice, Insurance Consulting and Technology, Willis Towers Watson.
“However, insurers are not maximizing their internal customer data utilization — they shouldn’t overlook what is readily available. They should also take it a step further and supplement internal data selectively with external data sets to improve results.”
Progress around the use of artificial intelligence and machine learning also fell short of expectations, though the percentage of respondents who use both to build risk models for decision making (26%) and reduce manual input (22%) doubled in the last two years.
Additionally, 69% characterized their current state of insurtech integration as early stage, while 22% said they’re not doing anything in this area.
“There are levers insurers can pull to sharpen their focus around implementing advanced analytics, such as reviewing staff allocation to get the most from employees who can devote sufficient time and develop their skills in the process,” Sukow continued.
“Overall, insurers should plot a course for how to apply data and analytics but remain flexible to take advantage of market and technical developments as they arise. Think of the whole process of enhancing analytics capability as a journey, not a destination.”
