Double-Edged Data in Insurance (Part 1 of 2)

Resilient
3 min readJun 4, 2019

--

How insurance providers can use consumer data for social good profitably

Underwriters have always sought to gather as much information on consumers as possible before insuring them. In the last few decades, their ability to collect data has been significantly enhanced. This increased access to data has its dangers and its opportunities.

Access to consumer data is easier than ever before

Tech innovations are giving insurers a host of new sources to gather consumer data. A few examples:

  • Consumer profiles and behaviors are digitized online on various websites, social media, etc.
  • Consumers are engaging with platforms that provide insurers with machine readable/processable information at scale.
  • With the increasing array of connected devices and Internet of Things, providers have greater potential to gather a more complete picture of consumer behaviors.

Along with this increased access to data comes a growing distrust among consumers regarding how companies are gathering, using, and protecting data. The many cases of cybersecurity breaches and data being stolen and/or misused have made consumers wary.

Nevertheless, insurers are continuously seeking new ways to get as much consumer data as possible, even looking to such sources as social media posts and personal Fitbit data.

How will (or, do) the biases in consumer data affect the way providers underwrite risk for the general population?

The double-edge of consumer data

Once underwriters have the data, how will they use it? To serve consumers better? Or to filter out all potentially unprofitable consumers, leaving behind large demographics?

Often biases in consumer data have led to the latter. Rather than create more targeted products, underwriters have left behind those considered too risky to protect, most especially those who are low- to moderate income (LMI).

However, “too risky” is not a strictly quantitative analysis. The interpretation of data, whether by human or machine, involves biases that can lead to misleading conclusions.

Consider LMI consumers, for example. A typical upmarket insurer would price the risk higher for a LMI group given the perception that their claims would be a bad financial risk. But the actual behavior of the LMI population reveals that this group tends to work even when they are ill, because they are often at the edge financially and cannot give up a paycheck. And so many will only claim to their insurers when they have absolutely no other options left. Given this behavior, suddenly this demographic could be considered a “good risk”.

Bias of interpretation have created disparities in coverage based upon hypotheses which may not be accurate.

This is the double-edge of data. The data we gather can either be used in service to our biases (which often lead to a disservice for consumers), or the data can be used to get a clearer, more detailed picture for consumers to create more targeted and relevant insurance products.

How consumer data can be used by insurers for social good

Profit margins are how underwriters have justified leaving behind LMI consumers. But it’s not as difficult as it’s often claimed to insure the un- or underinsured LMI demographic profitably.

Consumer data can create a highly detailed picture of behaviors and lifestyles to inform more relevant and targeted products, leading to:

  • better price points, benefiting provider and consumer;
  • improved delivery through the right channels; and
  • optimized workflows, reducing waste in processes and increasing provider savings.

In other words, the key is in how the data is used. Data should be gathered not to exclude consumers, but to better understand consumers to develop the right products, deliver them via the right channels, and incorporate the products in the right ecosystems.

Grameen Bank, Chime banking, and Petal, the credit card company, are just a few examples of financial companies using data in this way to create relevant products for often underserved consumers.

In Part 2 we dive into the detail of how insurers can create relevant and targeted insurance products using data.

Read part 2 here.

Learn how Resilient’s Inclusive Insurance programs can benefit your business. Visit imresilient.co, and contact us as Maximilian.Weiner@imresilient.co.

--

--

Resilient

Providing inclusive insurance to low- and moderate-income Americans