Social media and the insurance industry
A night on the town celebrating a friend’s 21st birthday. Cocktails. Shots. Even a nightcap. Exciting and memorable, right? But for an insurance risk manager, a dangerous mix.
And every year, insurance companies pay out millions on liquor liability claims.
Insurance risks increase with increased alcohol intake. One source states that to defend one claim is $150,000 and that a settlement in a liquor liability case could cost a business $500,000 or more. This creates real risk for insurers restaurants and other businesses.
The property and casualty insurance industry has remained fundamentally unchanged for over a century. Carriers compete based on the data they can collect and how they analyze it. What has changed is the type and amount of data available. Social media is different than traditional sources due to its ever-expanding nature and unstructured property. These characteristics offer tremendous opportunity but also foster skepticism around usefulness, privacy, and implementation challenges.
Cognizant SocialLense Introduced
Cognizant has pioneered Cognizant SocialLense, a business-consulting and technology framework that helps P&C insurance companies systematically gain social-media insights. Cognizant SocialLense translates information from consumer sites like Yelp and TripAdvisor into insights that can influence underwriter decision making.
The business relevance of Cognizant SocialLense hinges on two steps in the underwriting process: risk classification and the exercising of underwriter judgment. First, social media can confirm classification and alignment with risk appetite, such as if an establishment is a bar versus a restaurant or if alcohol is served. Second, Cognizant SocialLense can mine data to adjust pricing based on what people are saying about specific risk dimensions.
At the heart of Cognizant SocialLense is a taxonomy for contextualizing structured and unstructured social-media data to the underwriting process. Cognizant SocialLense comes preconfigured with social-media data points, insurance perils, and pricing attributes. With this as a starting point, the framework can be modified for the needs of a given book of business or set of objectives.
How to test the waters
To demonstrate how Cognizant SocialLense enhances underwriter decision making, we will typically progress through three steps with a given business unit or client:
- Business matching: The first reality check is to confirm that a material portion of policies can be matched against social-media sources with confidence.
- We have a fuzzy matching methodology and outcomes stratification tailored to the challenges of matching insurance-company data against sites like Yelp for small businesses.
- Based on our experience, match rates can range from 45% for a diversified book of business to 80+% for restaurants.
- Rules customization: After ensuring a satisfactory match rate, we focus on building rules to kick out accounts for manual review based on the propensity for misclassification and concentration of high-risk attributes; our taxonomy framework comes with a preconfigured rule set which is then further customized to minimize false positives and better align with underwriting goals.
- Process redesign: The final step is to evaluate and redesign, as required, possible changes to the underwriting process itself; some examples could include a feedback loop to agents to clarify classification incongruities or the manual review steps for an account triggering high-risk social-media indicators.
Cognizant SocialLense is a business and technical construct for viewing social-media data through a risk-management lens. Incorporating social media into the underwriting process is on the horizon of strategic initiatives for several of our leading commercial insurance clients, and in the coming years we expect the application of such insights to mature from exploratory to differentiating to must-have.
Please contact me to initiate a conversation on how Cognizant can help you navigate this journey.