Historically the profitability of most property/casualty insurers has been heavily predicated on policyholders’ future behaviors, ranging from the propensity to become involved in an accident to the likelihood of renewing a policy. But it’s only recently that researchers have begun to realize that social media can play a significant role in forecasting the actions of individuals. A recent study from the University of Cambridge found that in some cases individuals’ Facebook “likes” may more accurately predict future behaviors than a professionally designed personality questionnaire. With more than 40 percent of Americans now engaging with Facebook daily (according to recent Pew research), insurers might want to consider how they can use data from social media to tailor offerings to prospective policyholders’ “likes” and preferences.
Current Uses: Brand Perceptions and Claim Adjusting
Many property/casualty insurers are already aware of many of the possibilities of social media data. Initial insurer use cases focused on “listening” for brand mentions and other social media measures that indicated how insurers’ products were perceived in the market compared to their competitors.
More recent implementations have integrated social media data more deeply into decision making. For instance, some claims adjusters may use online dashboards to view information that claimants have made publicly available on social platforms. That information can include basic facts surrounding an event, such as a photo of a torn sail that led to a windsurfing accident. It can also include potential inconsistencies in the claim, if the claimant “tweeted” a windsurfing picture following an allegedly debilitating injury. Such capabilities are increasingly being extended to insurer underwriting processes to help agents identify cross-selling or savings opportunities and alternatively to spot potential inconsistencies on policy applications.
Potential Uses: Predicting Policy Longevity
Existing property/casualty use cases have only scratched the surface of what’s possible when you apply predictive analytics to social media data. In life insurance, consumer data, including online shopping habits and leisure activities, may predict policyholder longevity nearly as accurately as blood tests.
Social media information may prove equally effective in predicting policy longevity. A recent survey of 30 insurer CIOs identified bolstering growth and retention as the most important influence on their 2015 IT plans. Social media data is uniquely suited for these purposes: It’s timely, publicly available (often through open Application Programming Interface, or APIs), and contains a variety of information, including past consumer experiences (for example, product or restaurant reviews). In an insurance setting, Cambridge’s ability to predict personalities using “likes” could be applied to predict anything from policyholders’ “agreeability” to a renewal or cross-sell to “conscientiousness” in risky situations. Insurers that use social media analytical tools intelligently and responsibly can likely create a higher value and more meaningful repartee with each policyholder.
Participate in Our Social Media Analytics and Insurance Survey
The early movers in social media data analytics will likely be poised to acquire a competitive advantage as did the insurers that first explored insurance scoring or usage-based insurance. At ISO, we’re developing tools to help insurers incorporate social media data into their decision making more effectively. These tools may include strengthening brand analytics with our property/casualty domain expertise and creating “model-ready” data from social media for loss and customer lifetime value analytics.
If you’d like to learn more about social media analytics in insurance, please contact me at JWeiss@iso.com or 201-469-2216.