New York-based homeowners and renter insurance startup Lemonade recently released the second of three planned installments of its “Transparency Chronicles,” as the company continues to offer what it calls “full frontal transparency” when it comes to keeping anyone interested informed about how it is functioning from an insurance perspective.
Last fall, Lemonade Co-Founder Shai Wininger took the unusual step of publicly sharing metrics and sales data from the company’s first 48 hours. He did so again on Jan. 18 with the first installment of the Transparency Chronicles in what was called the “Lemonade Startup Report.” It features “interesting metrics” from the company’s first quarter in market — Q4 of 2016.
The middle installment, with highlights featured below, is written by Lemonade Chief Underwriting Officer John S. Peters. It includes loss ratios, risk distribution and licensing progress. The final installment to be released soon will be the Lemonade Social Impact Report, written by Chief Behavioral Officer Dan Ariely.
Here are a few highlights taken straight from Peters’ report (prefaced with Peters’ admitting that 100 days of data isn’t that meaningful in insurance and doesn’t establish a predictable trend):
• Our customers are new to insurance
More and more people are switching to Lemonade every day, coming from very well-known insurance companies. At the very least, this is a good early sign. But, more importantly, we’re bringing in a new breed of customers – the underserved. The majority of customers insured with Lemonade are actually new to insurance.
• Our customers are diverse
In insurance, you never want just one type of customer. Since our goal is to share risk across customers, too much concentration isn’t ideal. While we have a lot of renters, almost 50% of our premium came from homeowners – many with over $500K and up to $1.5M of coverage.
Our customers are geographically diverse as well. While 35% of our premium is from Manhattan, the rest is spread across the NY metropolitan area, and across the state.
• Our customers represent ‘high quality’ risks
Perhaps the most important part of underwriting is ensuring we don’t have ‘adverse selection.’ This means insuring someone only because our prices are (too) low, or because we attract the riskiest policies. We use a series of factors to help predict if a risk is better or worse than average, and to price it as accurately as possible.
Two bits of good news in this regard:
1. Potential customers that come to Lemonade for insurance score above average as ‘risks.’
2. Even more important, the best risks tend to end up buying Lemonade!
Having said that, there’s a lot we can improve on the underwriting side, and we’re working hard to train our algorithms to make better decisions. The more data we gather, the better our algorithms get.
• Our customers are helping us to make insurance into a social good
We celebrate claims when they come in. We’ve had a few (6 in 2016, to be precise) – exactly the number we expected based on industry statistics. What is different is that our claims have all been small – way smaller than industry averages. It’s too early to mean much, but right now our loss ratio is much better than the rest of the industry, and I’m hoping this hints that the Lemonade Giveback will be strong this year.
• The Nitty Gritty Insurance Numbers
Right now, over 25% of people that get a price, buy. This is high by any standard, both in insurance and tech. What is extraordinary, however, is the trend. The percentage of people buying when they get a quote is increasing every single month. In December, we were up to 36% for renters, and 26% overall.
Our “written premium” (basically how much insurance we sold) in 2016 (the last 100 days of the year, really) was $179,855. Our “gross loss ratio” (or claims we received in 2016 divided by “earned” premium) was 20%; a portion of that should be recovered from another insurance company, so we expect our final loss ratio for 2016 to end up at about 12%. While our reinsurers are standing by to help pay losses, we have not needed them yet.
• What We Need to Work On
While it all sounds great, some things did not work as we expected, and we had to adjust accordingly:
1. Get More Data In:
There is a ton of information out there that can help sign customers up faster and ensure the coverage is right. While we gather and implement a lot of data, we’re only scratching the surface. For some homes, we need to ask you the square footage, which is kind of lame in this day and age. We’ve made it a priority to seek out and incorporate new and different data every day to make sure policies are appropriate, and our customers are protected.
2. Innovating In… Language!
When we started selling policies back in September, we wanted to make sure regulators and customers were comfortable, so we launched Lemonade with the industry standard insurance policy contract. We know that it is poorly written, and unless you have a law degree – frankly even if you have a law degree – it can become confusing.
Who knew your liability coverage actually changes depending on whether or not you are also an insured under a policy written by the Nuclear Insurance Association of Canada (Section II.F.5.a.(3)). I assume most of our policyholders do not. We’re going to improve the policy so you can actually read it and understand what is covered and what is not. It will take time, but we will do it – I promise.
3. Improving Our Coverages
Our policy is great for most people, but isn’t as customizable as we want it to be. For example, at launch we could not add fine art – now we can. Need your landlord listed on the policy? We added that a few weeks ago. Identity theft coverage? Still no, but that will be here in a week or two. Kidnap and Ransom coverage – that one is a little farther off. Kudos to our customers, whose patience is crucial as we continue to build a policy that covers everything you want to protect. If we can’t protect it yet, we will tell you… and know we are adding options every day.
More on Lemonade:
