Social Media, Life Insurance and Behavioral Science

Micro targeted digital marketing and integrated data models will lead to fundamental changes in the life insurance industry.

Audience targeting in display and social marketing allows insurers to target customers with a particular risk profile. The idea of marketing to prospects with a favorable risk profile is not new, for example, carriers have pursued distribution via professional organizations whose members display lower risk for arguably since the development of guild based benevolent societies in ancient Rome. However, recent developments in the scalability and precision of  digital marketing technologies and the emergence of startups with integrated marketing and core operations data models has begun a fundamental shift for the insurance industry.

If you use social media and are “relatively young and healthy” you’ve no doubt seen the ads promising what appears to be an extraordinary deal on life insurance. Most often, the ad copy says something like “run 5 miles a day? Get $1 m in life cover for $x”.

The companies behind these ads are turning the life insurance industry upside down and although small players in the industry, rapidly gaining traction. By way of examples, Ethos Life, Ladder Life and Health IQ are backed by investors including Sequoia Capital and Andreessen Horowitz and have raised close to $100 m in capital each. Growth rates quoted by some of these companies include “400% in 4 months”.

Without knowing the details of these private companies’ business models, and observing that some make more use of social media and display marketing than others, the ads did make us think about what marketing insurance on Facebook really means.

As any marketer knows, the keys to success on Facebook are engagement and targeting. When executed well, both of these factors drive down customer acquisition costs whilst increasing the value of a customer viewing your ad to you relative to competing advertisers. What’s less well understood is the value of display marketing audience targeting factors in predicting losses. In insurance, carriers create a competitive advantage by gathering data from rating and underwriting systems and correlating these with losses. This “moat” is built up over a long period of time at the cost of paying claims.

Audience targeting factors like age and gender are relatively straightforward and their implications for rating and underwriting are well understood. Other factors however are more opaque. For example targeting people over the age of 18 that are “interested in health & wellness” on Facebook yields a target audience of approximately 50 million in the United States.

What this audience comprises exactly is determined by Facebook’s algorithm and since it’s part of their “secret sauce”, we’ll never know exactly what factors determine eligibility for the group. However, the new generation of Insurtech startups are acquiring customers in vast numbers on these platforms and are, in our opinion, building a “data moat” in the process which once established, may be hard to compete with. Whilst most of these companies are acting as intermediaries for established carriers, recent trends suggest that the broker business model is often a step on the path to risk participation.

Said differently, these companies are in a better position to know what the influence of being part of the “interested in health and wellness group” is on future claims and profitability. Of course there are regulatory hurdles in terms of what can be used for rating and the credibility of the data, but with the amount of data being generated through these platforms, it may not be long before some of these factors find use. This could be in the form of merely bidding more aggressively on audiences known to be more profitable.

It’s important to understand the blurring of the lines between marketing and product development that’s taking place here. Companies with integrated, end-to-end data models are creating a substantial advantage in terms of understanding how factors that drive both acquisition costs and comprehensive lifetime value interrelate. We call this “life-cycle marketing” and will take a deeper dive into it in our next post.

Another emerging social media marketing trend we’ve seen concerns customer behavior modelling. Marketing technology based on behavioral science is being used to acquire customers at costs orders of magnitude lower than available through traditional and search channels.

This technology actively engages users on social media platforms in “conversations” regarding topics of interest to the user. The data emanating from these AI driven interactions is being used to create behavioral profiles, which are used to inform the most effective method and messaging with which to further engage a potential customer. 

The resulting targeting and engagement can be powerful. A Swiss-Israeli startup, BestFit Human Intelligence, leverages this technology, based on the research of Professor, Dr Moran Cerf, for clients in the insurance industry achieving conversion rates between 20% and 30% in cross selling applications.

That we are living in times of immense change in the insurance industry is an understatement. The integration of new sources of data into product development and marketing models is already a strategic priority for many insurers. We believe however that a change in perspective can reveal that some of the most valuable data is already publicly available or readily obtainable.