A New York startup shakes up the insurance business

IT IS not typhoons or earthquakes that insurers should fear most, but geeks alert to their businesses’ inefficiencies. Daniel Schreiber and Shai Wininger, tech entrepreneurs with no insurance background, spotted that the industry is huge (worth $4.6trn in global premium income a year, reckons Swiss Re, a reinsurer), distrusted, antiquated and hopelessly unreformed.

In September they started Lemonade, a New York-based insurer for homeowners and renters. Some describe it as a peer-to-peer insurer (“Spiritually we’re a tech company,” says Mr Schreiber). Most agree that its app makes insurance a lot easier. This appeals to the digital generation: of 2,000 policies sold in its first 100 days, over 80% were to first-time buyers.

Insurance, the founders reasoned, suffers from misaligned incentives. Every dollar paid out comes from insurers’ pockets, encouraging poor behaviour. Normally upright people have few qualms about defrauding their insurer (as 25% of Americans do), pushing up premiums. Lemonade’s solution is to take 20% of premiums as a fee and to reward under-claiming customers by giving a share of unused income to a chosen charity.

This brings good publicity. But just as important is how different Lemonade looks behind the scenes. Instead of underwriters it uses algorithms; and instead of expensive brokers and salespeople it uses chatbots. It even uses AI and machine-learning to handle claims, a job typically seen as needing a human touch.

Late last year a customer called Brandon claimed for a stolen coat. He answered a few questions on the app and recorded a report on his iPhone. Three seconds later his claim was paid—a world record, says Lemonade. In those three seconds “A.I. Jim”, the firm’s claims bot, reviewed the claim, cross-checked it with the policy, ran 18 anti-fraud algorithms, approved it, sent payment instructions to the bank and informed Brandon. The real-life Jim (Hageman), Lemonade’s chief claims officer, was driving home for Christmas at the time.

Lemonade’s bots are still learning and pass more complex claims to humans. It is hoped that one day they will handle 90% of claims. In an industry with expense ratios as high as 30% this could offer huge savings. But there are limits to the claims that bots can be let loose on. And insurance dinosaurs have one advantage: data. For bots to get really clever they need lots. If Lemonade’s customer numbers remain small, they will not learn fast enough to stay ahead of the big boys using the same technology.

But insurance moves slowly. Miguel Ortiz from BCG, a consultancy, says that the big bet for Lemonade is that “it can stay ahead of a sleepy industry by doing standard insurance processes better than everyone else.” Already, the shake-up it promises has added some fizz and zest.