David Schapiro (DS): Could you please tell us about yourself, and how you decided to
enter the insurance industry and founded Kin?
Sean Harper (SH): I was a nerdy kid and learned how to program at an early age. That flowed naturally into entrepreneurship because once I had made some software, I wanted people to use it, and I wanted to make money from it, so I had to learn about marketing and business. For college, I went to the University of Chicago, where I became very interested in finance. That combination
of interests – finance and technology – naturally led me to fintech.
Financial products are inherently virtual goods, so banks, insurance companies, etc. are basically software companies in disguise with their entire business being conducted on a computer. However, most are not particularly good at software, which creates the opportunity for guys like me and my cofounder Lucas, who has a similar background.
Immediately before starting Kin, I started a payments processing business very similar to Stripe. We had a good developer API and the ability to underwrite a customer in real-time, which at the time was a unique capability. I sold that company and was looking for something new.
We were looking for a financial product that met three criteria:
1. Large and homogeneous market. We wanted to build a business that could get huge in its initial market, not have to piece together lots of niches.
2. Low competitive bar and overly intermediated. By going direct-to-consumer, we could undercut on price, maintain high margins, and delight customers, who are less and less interested in intermediaries.
3. New data sources available for pricing and underwriting. We knew we wanted to win on risk selection and could do so with the right data.
Homeowners insurance met those criteria very well:
1. $105 billion market and growing, especially as houses get more expensive and the weather gets worse due to global warming.
2. Expense ratios have not improved at all in the last 30 years, while the prices of other financial services have dropped considerably. That’s in part because 93 percent of homeowners insurance is still sold through retail agents.
3. There is a lot of data about buildings that didn’t exist before, and what you see in a public search engine is the tip of the iceberg. Meanwhile, legacy insurers remain dependent on the user- and agent-reported data, which is incomplete, inaccurate, and biased.
DS: As someone who has been looking from the outside in and is now part of the insurance industry, what do you see as some of the key complexities in the insurance supply/value chain?
SH: There’s a lot of work being done by humans in insurance that could be done by machines.
If you spend the day in an insurance agent’s office, you will see that the agents spend very little time actually speaking with customers. They spend a lot more time wrestling with the carriers’ archaic IT and processing paperwork. Inside the insurance company is similar with lots of human involvement in underwriting and processing.
Those are tasks that a machine can do more efficiently than a human. Some of our peers have focused their AI efforts on trying to eliminate customer service, which I think is a big mistake because humans like talking to other humans. We have instead focused our AI efforts on eliminating boring and repetitive back-office tasks, of which insurance has many.
In the rest of financial services, there has been a big trend of decoupling the origination and servicing (basically the customer interface) from the capital source. That way, risk appetites can change and risk can be traded without disrupting the customer. That has not happened yet in insurance, but it’s starting to as balance-sheet-light carriers like us continue to grow and as the
reinsurance markets become increasingly analytical.
I think a lot of legacy carriers will reach an existential crisis: what are these companies? Are they customer acquisition companies? (Not really – they outsource that to agents.) Are they origination and servicing companies? (Not very good ones because that is primarily a technology business, and these companies are not quick to adopt new technology.)
But just being a balance sheet is a pretty commoditized business, especially when capital from outside the insurance industry can compete with insurance capital as with insurance-linked security. It’s also a scale-intensive business, where size and diversification pay off big, and all of the American primary carriers are dwarfed by international reinsurers.
DS: Starting from the customer side when purchasing a policy, what are some of the
challenges you see in the distribution channels that are addressed when becoming vertically aligned?
SH: Customers think it’s quite strange that they cannot just get a quote and buy home insurance directly. It’s also confusing from a servicing and claims experience because it’s not clear what is the responsibility of the agent vs. the carrier. Because there is not a direct relationship between most insurance companies and their customers, the companies do not feel the same responsibility to create an enduring relationship with the customer. Kin is different. We have a direct relationship with all our customers, and we treat them like our family, which is why we decided to name the company Kin. It literally means “family.”
DS: And if we jump to the other end of the customer experience – How do you see the
claims handling process and level of service?
SH: Claims is one of the most important customer experiences in insurance. It’s the only time customers really engage with the product. But today, many insurers / MGAs (managing general agents) outsource their claims handling to third parties, which means they’re also outsourcing one of their opportunities to differentiate their service to customers. By owning the claims process, we can
control the experience and keep improving. There is also a huge opportunity for technology in claims, especially in extreme weather situations. Lots of claims happen at the same time, and using technology to streamline that can vastly improve outcomes.
For example, Kin has a bot that texts customers to check on them when they are hit by extreme weather. The customers love it, and when there is damage to the house, we can get the claim sorted out that much faster.
DS: Do you see ways of improving the cost of service to the insured in a vertical model?
SH: There are so many handoffs in the legacy value chain; at every handoff there is waste. For example, in the legacy model, the agent enters the information about the policy and then an underwriter checks it, which is redundant.
The agent model is also inefficient. A carrier is doing a decent job to achieve 5 percent net margins but is also paying out 15 to 20 percent to the legacy agents and the infrastructure to support them. It’s not a particularly high quality experience, either. They are not quality controlled, and they spend more of their time with the overhead inherent in running a very small business and wrestling with the carrier’s outdated technology than actually helping customers with insurance.
DS: And how about reducing the time it takes to respond/service a customer request
SH: As I mentioned earlier, we are taking claims via text message, can get a photo right away, and have a preliminary estimate very fast.
With the legacy insurers, claims are handled offline, so it’s hard to even know the status of the claim or what happens next. I can track my delivery pizza getting made, but not my roof replacement.
DS: Off the cuff, by focusing on the customer and distribution side (and not the underwriting), it seems that an MGA model could provide better customer service. What are some of the challenges in the MGA model and the strategic benefits a full-stack (vertically integrated) insurer has?
SH: There are no large MGAs, so it seems safe to conclude that the MGA is not a scalable model. In some other areas of financial services — such as lending and payments — there is a well-developed rent-a-charter ecosystem because starting a bank is nearly impossible. In insurance, there is a concept of fronting and MGAs, which is similar to rent-a-charter but not nearly as well developed. It is also relatively easier to start an insurance company than a bank, although it is still quite hard.
At Kin, we are building what will become a big, iconic company, and we cannot put our fate in the hands of a fronting company or reinsurer who does not have that same commitment and whose priorities could change at any time; it is too large an existential risk. Furthermore, we are building a revolutionary company, which is predicated on doing things differently than the legacy industry.
Therefore, we cannot have legacy players validating our day-to-day decisions through the legacy lens.
DS: The cost of acquiring a customer (CAC) is inherently high in insurance. How does vertical integration reduce these costs?
SH: One advantage we have is that we can directly market only to the customers who are a good fit and who are likely to convert. The legacy industry isn’t as good at performance marketing like many other areas of financial services (for example credit cards), and basically just pays agents and does branding. In both cases, they are spending a lot of money attracting customers who are not necessarily a fit.
I think a big reason why CAC is so high in the legacy industry is also because it’s such an undifferentiated product. Our product is better, and we continue to widen that gap, which results in more customers seeking us out organically, which reduces CAC a lot. If we look to consumer products, many of the really successful new brands have a product that’s better or at least different, even in things that have historically been commodities, like Warby Parker or Harry’s.
DS: Could you please provide some parting words that you would like to share with other insurance CEOs?
SH: Embrace the hard problems. Solving them gives you more competitive advantages than if you only incrementally improved an established process.