Behavioral Insurance

Twenty years ago I felt very alone as a behavioral person working in strategy/business development of a leading Swedish insurance company.

I was asked to help produce a new strategy for the corporate insurance division, so I looked at the old strategy and the customer base. There was no indication that the sales force had paid even the slightest attention to the overall strategy in their work. So I looked at the previous strategy, and the one before that. Still no sign. No matter which industry my company had decided to focus on, the sales people had just been going about their business as before.

Finally, ten years back, there had been a strategy that I could actually see some traces of.  The freshly recruited president had spent several months piggybacking on employees in different offices before taking up his new job, and then he had launched the Biblically inspired idea that “we insure those that take care protect themselves”. Ten years later, we had a clear over representation of financially healthy companies on our customer list. Further analysis showed that fast growing companies have more claims. The slowest growing companies also had more claims. The most profitable insurance business was in between.

We didn't want the Gazelles

We didn’t want the Gazelles

Instead of looking at any specific industry, we decided this would be our new strategy. The beautiful thing was that our competitors were all looking at different industries. This meant we could focus on financially healthy companies with medium growth for many years before they could even see what we were doing.  In an oligopoly that seldom happens since every customer we take is a customer lost for one of our four competitors. That means a concerted effort to take market share in e.g. aviation creates an insurarance price war in aviation, which quickly kills any profitability. They all looked at the market industry by industry. Being able to go below their radar was great for us. This meant we reached  a claims ratio around 50 (i.e. 50 cent per Euro of insurance premiums was paid out in claims) while our competitors were close to 100.  Another good thing was that we were really just operationalizing what our local sales people had already been doing for years. They loved hearing that!

I figured past behavior is the best indicator of future behavior. So I asked an actuary to run an analysis on companies that have had no claims for two years. Sure enough. They were the backbone of our business.  Next I asked a retired insurance claims manager to spend three months costing a thousand denied insurance claims from companies that previously had had two claim free years. That analysis made it possible to construct a one page insurance policy.

Read your insurance policy. I bet it’s a 10-50 page booklet full of sentences all starting with something like “this insurance does not cover….” and then there is a list of everything from earthquakes to avalanches, terrorist attacks, bugs and tsunamis. All of this is either the result of a negotiation that probably took place long ago at Lloyd’s of London or the result of an analysis of claims.  There is usually a good reason for most of these policies to be included somewhere in the world, but none of them are valid everywhere. They are the auto pilot of the insurance industry.  We were mostly going to take away limitations that are as likely as Sweden being attacked by Nepal (note to American readers: Nepal and Sweden are two small nations in different parts of the world). A one page insurance policy is a much better product than a 50 page insurance policy where every sentence limits the scope of the policy. We wanted to sell good sleep, and we figured we would never have to engage in another price war to keep these great customers.

one page insurance claim

After a long debate the management group declined this project. That is a pity, as I think it would have had a major impact on the insurance industry.  Past behavior predicts future behavior and that makes it possible to do great things for your best customers.

My next project was in construction. This business was not very profitable. I had two students run simple, standard engagement surveys at 37 small and medium sized construction companies that we insured. Then we ran the engagement scores against the claims of these companies. The results were as we expected. Where engagement was high and stress was low we had highly profitable business. Where the opposite was true, so was the insurance result. Stress caused accidents and lack of engagement caused theft and other losses.

37 companies is not a grand study but the scale of these patterns was astounding. We basically paid ten cents of claims per euro premium to the construction companies with the most engaged work force, and very roughly ten euros per euro insurance premium to the worst ones. That meant that unless we bring engagement and stress into our price models  we were in effect sponsoring bad management.

construction worker running

Again, this project was eventually declined. Although clearly relevant, actuaries still saw engagement as too intangible to be a price variable.  If the engagement scoring was to be done by sales people, there was a concern that they would rate every construction company as highly engaged in order to bring down the price and win more business. Underwriters and claims people were expected to do the opposite.

Now, all of this was twenty years ago, or more.  Now I’m publishing this to see if I can inspire someone to water one of these seeds. Please do, and help yourself to whatever fruit they may bring.


Tagged with: , ,
Posted in Uncategorized