September 19, 2017
Why don’t more people buy insurance?
by Paul Carroll
In the wake of Hurricane Harvey and now Hurricane Irma—and with Hurricane Maria now pummeling the poor Caribbean again while Hurricane Jose looks likely to lash Long Island—it’s becoming clear how much people have avoided buying flood insurance and how expensive their decision may be. The question is: Why don’t people buy insurance?
The answer seems to be that people often don’t think they need insurance and find it expensive. For good measure, people complain about the complexity of insurance and how unpleasant it can be to deal with insurers.
The problem isn’t just with flood insurance, either. Far from it. Life insurers, in particular, face dwindling interest, but the lack of demand for insurance is widespread. Some are now saying that flood insurance should be required, and I’ve heard others suggest that life insurance should be mandatory. Those ideas may or may not make sense—but shouldn’t the insurance industry aspire to producing products that people want to buy, not ones that, like auto and home insurance, they only buy when forced to do so?
That’s a bit of a roundabout way of saying I think insurtech can solve many of the broad problems facing insurance, slashing costs while reducing complexity and smoothing interactions with insurers. Exhibit A is chatbots.
The bots use artificial intelligence to answer routine questions, generally through some form of texting. Customers don’t need to sit on hold for minutes listening to sales pitches, be chastised because they don’t have all their documents in front of them, then get transferred twice because it’s not clear which department they should be calling. So, customers are happier. Insurers get to save gobs of money because a call center rep can handle five to 10 times as many customers as a rep can now. Insurers are increasingly using chatbots to simplify purchasing, not just service or claims processing—attacking what Brian Duperreault, now CEO of AIG, identified in an ITL article a year and a half ago as the “massive cost of doing business…[that] puts our industry at risk.”
An article in the past week by three senior partners at McKinsey identifies the sorts of efficiencies that are possible—a British broker that automatically processes 3,000 claims a day, managed by four employees; a department of 250 employees turned into 110 bots and 11 human supervisors; 160 bots that process 500,000 transactions a month. Remember, the changes make life simpler for customers and come at a time when the insurance industry faces a talent shortage.
At the risk of being repetitive: An article I highlighted last week provides a host of additional examples of how chatbots are ready to take on a huge role.
In addition, Pypestream, the company that we believe is the leader in chatbots in insurance, is announcing that it has received a round of financing from W.R. Berkley and will have a Berkley executive join the board. Pypestream, which raised $15 million in Series A funding in February, isn’t disclosing the amount that Berkley is investing. “We will use this additional investment to fuel growth across more industries and use cases, and to help more businesses provide 24/7/365 experiences that their customers not only love but expect,” said Richard Smullen, CEO of Pypestream.
Chatbots, alone, likely won’t be enough to get people to buy flood insurance, I’m sorry to report. But they can get us started on the sort of radical reduction in expense and improvement in simplicity that will lead the charge for insurtech. Perhaps the industry can even get to the point where people look to buy insurance, rather than only doing so when a regulator or mortgage bank forces them to do so.