Data Solutions
  • Articles
  • October 2019
  • 5 mines

Trick or Treat: How Smarter Segmentation and Data Use Can Help Insurers Win Policyholders and Avoid Ghoulish Risks

Trick and treaters looking for candy
In Brief
For some, there’s no spookier scenario than blood work and cumbersome forms to obtain insurance coverage. But as Neil Parkin of RGA South Africa explains, emerging risk segmentation techniques, combined with smarter use of data, could help ease these fears.

A friendly neighber kindly pours sugared delights into the outstretched hands, and the ensemble moves on to the next bell.

But vampires, fairies, and knights are astute creatures. They don’t ring every doorbell. They have a complex algorithm that helps increase the likelihood of a treat, and reduce the risk of tricks. By considering historical experience at each house, perhaps the toys in the garden or the kindly face peeping out from the window, these candy-laden visitors judge the targets with the best potential… and those to avoid altogether.

Like these children, insurers are searching the neighborhood for treats in the form of policies that can be accurately priced. Occasionally they receive tricks: a client’s risk level is far higher than expected or anti-selection and non-disclosure lead to an early claim.

While medical underwriting plays a key role in mitigating this risk, the cost and invasive tests required to complete full underwriting can deter many purchasers. Insurers face growing market pressure to simplify the underwriting process – both to enhance the client experience, and also to facilitate real-time online purchases.

To meet this consumer demand and also reduce the probability of getting a trick, insurers are increasingly turning to new and innovative sources of information.

Alternative Rating Factors

While underwriting and pricing utilize traditional rating factors such as age and gender, a rich set of non-traditional factors may provide a deeper understanding of risk.

One intriguing factor is a person’s social connection and support. Various studies show the potential predictive power of biological, psychological, and social conditions to influence an individual's resilience after a devastating injury or disabling illness. Moving from a purely medical to a bio-psycho-social model could provide a more holistic view of each applicant or insured, and thus potentially deliver a better gauge of the true risk to the insurer over time. 

Insurers are also increasingly looking to lifestyle factors which could ultimately help create dynamic models that change over time. Examples include sleep and physical activity.

Alternative Data Sources

The core challenge presented by alternative rating factors is measurement. Predictive analytics, combined with new technology (e.g., wearables) and third-party databases, offers a potential solution.

As an example, RGA helped validate a credit-based insurance score that models credit-based behavior and is highly predictive of mortality and lapse risk so that insurers can make additional (and more competitive) offers to qualified applicants.

In the United States, data from prescription drug histories and traffic databases has been harnessed to reduce medical testing, accelerating the sales process while enabling carriers to better manage risk. As South Africa moves toward a system of traffic demerits, or even a form of national health insurance, data sources could emerge in the future to achieve similar results. 

By obtaining reliable, more holistic data about potential clients, insurers could convert tricks into treats by segmenting risk more intelligently. This could create a win-win scenario whereby consumers obtain cover for a fair premium. For example, data around a diabetic's treatment adherence could enable an insurer to offer more affordable terms. 

While customer-centricity may be an overused buzzword, many application processes remain scary for those looking to purchase insurance. When insurers ring doorbells, potential policyholders often do not see cute fairies and brave knights; instead they see ghoulish ghosts waiving long application forms and needles, while trying to explain highly complex products.

How can the industry swap these scary costumes for more approachable attire? Insurers are working hard in a number of areas:

  • Creating digital interfaces: Although the need for financial advice means that most comprehensive insurance policies are not sold online, customers still value the ability to obtain information digitally. Further, customers increasingly want to manage individual policies or engage electronically with insurers.

    Digital interfaces can’t just be slick websites – insurers need to meet customers where they are, including social media platforms.

    And it’s not enough to attract a customer with a friendly mask, only for them to discover a monster behind it. For example, some insurers still present "online" forms on a website that must be printed, completed, and then scanned to send to the carrier. 

    Simplifying the application process starts with improving online interactions with each customer, but it doesn't end there. Carriers need to simplify and accelerate the back-end processes that support the business throughout the policy lifecycle.

  • Applying behavioural science: Studies suggest small and low-tech adjustments based on behavioural science insights can reap great rewards. The phrasing and ordering of questions on application forms can significantly improve the information insurers receive. For example, RGA tested re-wording a smoking declaration requiring a simple “yes-no” response to a question with a range of options. In this experiment, smoking disclosure rates increased from 35% to 52%. 

    By moving away from the notion that people are rational, life insurers may make greater inroads into an ongoing challenge: offering a great product that people need, but don’t buy.

  • Enhancing the customer experience: Although many insurers are making great strides in accelerating and simplifying underwriting, other process enhancements may also be worth consideration. For example, insurers can:
    • Use data (e.g. to pre-populate forms) and make other application design enhancements to make the process as efficient as possible,
    • Help people complete medical forms through tele-underwriting and more intuitive design,
    • Automatically set up and manage appointments at the nearest testing facility on behalf of the applicant through calendar apps and bots,
    • Embrace dynamic underwriting models that adapt to new data after initial underwriting is complete, empowering the insurer to stay in communication and provide attractive new offers to the policyholder. 

There are risks in over-simplifying underwriting risk assessment. Insurers must balance rising anti-selection exposure from weaker underwriting rules with fewer policies issued through more strict underwriting rules. In addition, non-disclosure rates or claims disputes could increase as the simplification of questions leads to reduced medical clarity. The power of data-based rating factors must also be balanced against the need for privacy and compliance with regulatory regimes. 

As an industry, insurance provides valuable financial support to people when they need it most. Insurers enable the education of children, protect the assets for those left behind after a breadwinner’s death, support the critically ill, pay the income of those unable to work, and help ensure a dignified goodbye to loved ones.

Perhaps then, insurance is not best associated with Halloween. Dare we suggest that insurance is truly an act of love, and worth being celebrated instead on Valentine's Day?

More Like This...

Meet the Authors & Experts

Neil Parkin
Neil Parkin
Head of Business Development, RGA South Africa