Counting the Cost of Cancellations: The Power of Data in Predicting Insurance Cancellations

There is no denying that mid-term insurance cancellations cost money and are a particular challenge for the broker community, already operating on slim margins in the motor insurance market. By the time a customer decides to cancel their policy, the insurance provider has incurred marketing and admin costs, as well as aggregator fees in most cases.

Our research at LexisNexis Risk Solutions has shown that each cancellation costs an insurance provider between £25 and £75. For an insurance provider with 100,000 policies on their books, a cancellation rate of around the industry average of 5% per year, equates to a loss of anything from £125,000 to £375,000 each year. A significant sum for any business.

Much of this cannot be recovered from the customer. In addition to costs incurred, there is the potential loss of the lifetime value of that customer who could have stayed with the insurance provider for many years.  So the revenue is lost not only on this first policy, but also on future potential renewals as well as cross selling opportunities.

Whether a cancellation occurs within the cooling off period, half way through the policy, or as a result of a new business renewal, cancellations are a significant problem for insurance providers operating in the highly competitive motor insurance market.

To tackle this problem, we set out to examine whether data on past cancellations could be used in understanding future cancellation risk.

The scale of the insurance cancellation problem

A detailed study of the LexisNexis® Motor Policy History database, our first contributory database in the UK clearly revealed how past behaviour correlates to future risk. The analysis confirmed that not only are previous policy cancellations an indicator of increased potential for future cancellations, but they can also indicate a higher risk of claims and fraud.

The primary aim of the study was not to identify fraud indicators, but it quickly became clear that previous cancellations have a significant correlation with increased potential for fraud.

What we found is that between 10% and 30% of the customers currently on insurance providers’ books have had a cancellation in the past, and therefore present a higher risk of future cancellation, high claims cost and fraud.

Some 5 million current policyholders have cancelled an insurance policy within the past five years, 1.4 million have cancelled two policies.

But perhaps most striking was that in the last year alone, there were 1.3 million new business cancellations, almost a third of the 4 million new policies sold.

Expectations that most cancellations would be within the cooling-off period were quickly proved incorrect during the study. Only 15% were cancelled in the first two weeks, 37% within 16-100 days and a substantial 48% 101-364 days after purchasing the policy.

Insurance policy cancellation relativity

Detailed retrospective analysis of policies in the LexisNexis database has highlighted other factors which increase the probability of future cancellations. The analysis showed that cancellation breeds cancellation. If an individual has ended a policy prematurely in the past, they are more likely to do so in future. And the more policies they have cancelled, the more likely they are to repeat the behaviour. These individuals present an average 70% higher loss cost than someone with no history of policy cancellation.

Those with CCJs (5% of current policyholders) are also more likely to cancel, as are those who have attempted fronting at point of quote, and those who left it until the last minute before purchasing a policy.

In our analysis, the risk of cancellation was 19%, more than double the average, when someone purchased a policy to start on the same day, presumably because they rushed the decision and later regretted their choice.

30% of policyholders have no policy history or presence on the electoral roll. Whilst this group may not represent a higher risk in terms of credit or claims, they have an increased risk of cancellations. What we have found is that the more information there is available on the individual, the lower the risk of cancellation.

Gaps in cover also indicate higher risk of cancellation: the more gaps between policies, the higher the risk of mid-term cancellation.

Mitigating the risk and cost of policy cancellations

Now that historical data analysis and insights have allowed us to isolate where the cancellation risk lies, the motor insurance providers have a real opportunity to reduce loss costs and increase market share.

By leveraging cancellations data, insurance providers will be able to make price adjustments to account for the risk each individual represents.

This starts with running a retro analysis of cancellations to allow insurance providers to uncover specific cancellation drivers for their business and their customer base. They can then apply the intelligence to future business, helping to make better decisions instantly. The result is more intelligent decision-making and pricing, creating a higher quality book of business, further increasing competitive advantage.

Follow the link to the LexisNexis Risk Solutions website to find out more about how we support insurance providers.