How big is “big data” for life and health insurers? What are insurers worldwide focused on when it comes to predictive analytics? How are they adopting it and what are their expectations for the future? We set out to find answers to these questions through our global predictive analytics survey just released to participants this month.

Our latest research study gathered feedback from 136 different participating companies in 23 countries grouped into eight regions: Australia/New Zealand, Europe, Latin America, Middle East/Cyprus, South Africa, Taiwan, United Kingdom/Ireland and the United States.

For the purpose of this research survey, we defined predictive analytics as “statistically rigorous techniques (beyond conventional actuarial experience analysis) that are applied to data to model or predict future outcomes.” Predictive analytics has broad potential applications across different insurance functions: It ranges from marketing and customer engagement, to pricing and risk management via underwriting and claim processing. Adopting predictive analytics in those functions requires a paradigm shift, management commitment, resources and talent.

While predictive analytics has been widely used by companies outside the insurance industry - for example, Google’s self-driving cars and online recommendation engines - its usage in life and health insurance is still regarded as an innovation.

Our survey results showed:

• 22% of the participating companies currently use predictive analytics.
• 32% plan to develop the capability over the next two years.
• 46%, however, have no immediate plans to use it in their business in the next two years.
• More companies use predictive analytics for sales and marketing support than for any other function, which aligns with where companies feel there is the greatest opportunity. Sales and marketing support includes acquisition, engagement and retention of customers, as well as evaluation of intermediaries.
• 40% of participants report sales and marketing could benefit the most from predictive analytics, followed by underwriting at 30%.
• The majority of participants who either have built models, or are in the process of building them, have done so using internal resources. Over one-third of these companies have engaged, or are planning to engage, external resources (such as vendors or reinsurers) to assist, particularly around underwriting.

We also found some variance among insurers using predictive analytics in different regions of the world. Insurers who currently use predictive analytics, however, are in the minority across all regions:

The most significant barriers for insurers that are not yet implementing predictive analytics included:

The power of predictive analytics is still to be determined. Many companies reported that their models meet their expectations. Still, the issue of obtaining data remains a challenge. Companies rely on traditional insurance data when building their models; however, obtaining historical data is a considerable undertaking for many.

Regardless of obstacles to overcome, it is apparent that insurance companies are seriously looking at the usefulness of predictive analytics. If plans stay on track, over the next two years the industry is set to experience considerable growth in this area and undergo significant changes. We’re continuing to monitor trends - and are focused on solutions that will help further the successful use of analytics and applications for insurance.

For those of you who are not on a Predictive Analytics team but are interested in learning more about how the pieces work together, this graphic displays our findings: