Tag Archives: algorithms

Lemonade App

The Lemonade app is a new way of insuring.

The company disclosed that it raised $34 million in new venture funding to assist in broadening its operations.  General Catalyst led the round with assistance from GV (formerly Google Ventures), Thrive Capital and Tusk Ventures.

This funding was in addition to $13 million in seed funding from Sequoia and Aleph.

This app is not only involved in the insuring of autos and household goods but is committed to social good. It is part of their business model.

Insurers should fear geeks alert to insurers’ business inefficiencies.

Daniel Schreiber and Shai Wininger, the founders of Lemonade, do not come from an insurance background. They realised that the industry is worth $4.6 trillion in global premiums.

These two guys regarded the current state of the insurance industry as frustrating and outdated.  Insurance is one of the least trusted sectors of the economy.  It ranks very low in customer satisfaction.

The insurance industry has a poor reputation for trust and customer service with its complex rules, fine print and lengthy processes.

The New York based start-up hired Dan Ariely as its chief behavioural officer.  Ariely is a professor of psychology and behavioural economics at Duke University.

In the first 48 hours after it opened for business in New York, Lemonade reported that it sold 142 policies.

They generated $14, 300 in gross written premiums.

Lemonade takes a flat fee for insurance and gives back money to causes which policy holders care about.

Shai Wininger believes that behavioral economics  together with Lemonade’s unique technology will decrease fraud.  It will also get rid of bureacracy and lower costs for its clients.

Lemonade employs 30 people and uses algorithms to process claims. The company says they can make decisions quicker than regular insurance companies. This can furthermore be done at a lower cost. The have a bot called Maya instead of brokers and an app instead of paperwork. Insurance can be done in seconds. Claims can be made in minutes. They set the world record for a claim paid in just 3 seconds.

The site has a feature which is called “switching”.  It allows users to cancel other policies they may have, receive a refund and purchase a new policy from Lemonade with a click of a button.  New customers came from State Farm, AllState, GEICO and Liberty Mutual.

When you sign up with Lemonade you are asked to pick a charity. At the end of each year, if you and other supporters of the same cause don’t make too many claims, then a portion of the money you have paid is then passed on to the chosen non-profit.

Last year a customer made a claim for a stolen coat. He answered a few questions on the app and recorded a report on his iPhone. 3 Seconds later his claim was paid. In these 3 seconds “A. I. Jim” the firm’s claim bot reviewed the claim, cross-checked it with the policy, ran 18 anti-fraud algorithms, approved it, sent payment instructions to the bank and hence informed the client.

Studies of EY or Accenture show that clients do not trust their insurance companies to pay claims. A new product such as Lemonade could cause customers to switch in large numbers to a product that they can trust.

The disruptive possibilities of Lemonade lies in their beautiful user interfaces, artificial intelligence and use of big data.

Insurance and technology

The insurance industry has always been slow to grasp the different facets of technology. The reason lies in the fact that the industry is reliant on historical data. They are therefore understandably averse to risk.

The insurance industry is thus being put under pressure by technology. Customers want to pay an individualised insurance premium which reflects the actual risk taken by the insurance company.

The cyber security market is growing at an alarming rate. There is also pressure to incorporate and manage the risk more effectively. Traditional insurers are however reluctant to make plausible offers to vulnerable clients.

Huge amounts of data are provided by technology. In addition, insurers spend vast amounts of money on claim processing as well as fraud detection. New algorithms are able to predict risk. They allow for vast automation in the underwriting process. Furthermore, contracts can be managed more efficiently. Examples are wearables for healthcare and GPS trackers for cars.

Only about a fifth of insurers claim that they want to fully automate the process. Nearly 80% of insurers say they are struggling or just getting started with automation.

Between eight and fifteen factors are taken into account when an underwriter issues a policy. This may be a motor, home or life policy. As many as 60 answers may be required from a life insurance applicant. A sophisticated automated system could supply vast amounts of data.

Policy writers are wary of underwriting risk based on technology which supplies no justification for pricing.  Furthermore, staffing will be reduced with increased use of underwriting automation. Risk assessment algorithms will become more reliable. Executives will have increasing confidence in them. Low-level underwriting will become cheaper and more consistent.