Data Quality Continues to Challenge Insurers

The top challenges among insurers investing in and employing predictive analytics are a lack of sufficient data and limited numbers of skilled modelers, according to a survey released by Earnix, a provider of pricing and customer analytics software for banks and insurers, and ISO, a source of information about property/casualty insurance risk, titled “2013 Insurance Predictive Modeling Survey.”

Data quality is an issue especially among large insurers (more than $1 billion in gross written premium), with 57 percent pointing to it as their most significant challenge. Among smaller insurers (less than $1 billion in gross written premium), data quality (30 percent) was second to a lack of observations/skills (42 percent).

These challenges are becoming more and more prevalent within the industry, with as many as 82 percent of respondents currently using predictive modeling in one or more lines of business, including personal auto (49 percent), homeowners (37 percent), commercial auto (32 percent) and commercial property (30 percent). The most common use of is for pricing, where 71 percent of respondents use predictive modeling either always or frequently.

But there appears to be a caveat among analytics’ widespread use. While the use of predictive analytics is pervasive throughout the insurance industry, larger insurers are more likely to make use of predictive modeling than smaller ones. In fact, all the respondents from insurers that write more than $1 billion in personal insurance use predictive modeling, compared with 69 percent of the smaller insurers that took part in the survey (writing less than $1 billion in personal insurance).

The report goes on to say that “the role of big data in modeling initiatives is predominantly a big company affair at this point.” Among insurers with more than $1 billion in gross written premium (GWP), 51 percent either currently use big data or are evaluating or implementing big data initiatives, compared with 30 percent of the insurers with less than $1 billion GWP.

Regardless of size, though, insurers acknowledge a variety of positive results. According to survey respondents, predictive analytics is enabling insurers to drive profitability (85 percent), reduce risk (55 percent), grow revenue (52 percent), and improve operational efficiency (39 percent).

In terms of expanding on successes found through data and analytics, using additional data attributes was the most promising avenue seen by survey respondents to increase the power and quality of models built today.

The survey also revealed that data projects universally require patience, with insurers spending considerable time on data preparation and deployment before and after actual modeling work. More than half of survey respondents (54 percent) spend more than three months on data extraction and preparation, and more than two-thirds of the respondents (69 percent) take more than three months to deploy new models.

Survey responses were collected online from 269 insurance professionals representing companies that sell personal and commercial coverage in Canada and the United States.

For a blog from Joe McKendrick discussing the rapid pace of advancing analytics technology and its importance to insurers, click here.

For reprint and licensing requests for this article, click here.