Podcast: Bringing insurance to "uninsurable" homeowners with Kin CEO Sean Harper

‍

‍

‍

‍

Show notes

Bios

Bill Cilluffo joined QED as a Special Advisor in the fall of 2014 and became a Partner in 2015. He is currently Head of Early Stage Investments after six years as Head of International, leading QED’s Investment teams in Latin America, Europe and Asia.

Prior to joining QED, Bill spent nearly 20 years at Capital One, spanning several roles and leading several businesses. He spent the first 6 years of his career leading Marketing, product development and credit policy for Capital One’s subprime credit card business; ultimately having overall P&L responsibility, and growing the business to become the most significant player in the market. He moved on to spend 2 years in various new business development roles, spanning the telecom, medical finance and small business finance industries. Bill spent 3 years as Deputy Chief Credit Officer for the bank, playing nearly every role there was to play in the central credit function, after helping build the department from scratch in 2002.

Bill then pivoted his career to general management, leading Capital One’s Canadian, and ultimately International businesses, over the course of 6 years. Profitability of the business grew significantly under Bill’s leadership, through new product and channel introductions, acquisitions, and significant cost take out. During Bill’s last 3 years at Capital One, he led its Co-Brand and Private Label credit card business, building the business nearly from scratch to one of the top few players in the US market, through a series of acquisitions, most notably including leading the acquisition and post-merger integration of HSBC’s US credit card business, which closed in May 2012.

Bill graduated with a BA in economics from the University of Michigan, and competed the SEP program at Stanford GSB.

‍

Bill Cillufo:

You're listening to the FinTech Thought Leaders Podcast from QED Investors, your deep dive into the world of venture capital and financial services with today's digital disruptors. QED is a global venture capital firm focused on investing in FinTech companies all the way from pre-seed to IPO. FinTech Thought Leaders brings together the most talented entrepreneurs tackling today's biggest problems. If you're looking to learn more about what motivates our founders and team members to succeed, you're in the right place.

‍

Hello and welcome to the FinTech Thought Leaders Podcast. I'm Bill Cilluffo, head of Early Stage Investments at QED investors. Today on the podcast, I'm excited to be joined by Sean Harper, CEO and co-founder of Kin Insurance. Sean, welcome to the podcast.

‍

Sean Harper:

Hey, good morning, Bill. Thank you.

‍

Bill Cillufo:

Hey, just as a way to kick things off so the listeners have some context, I wonder if you can give us maybe a 6O second pitch on what Kin Insurance is and what you guys do.

‍

Sean Harper:

Yeah, of course. Kin is a high-tech provider of homeowners' insurance. Homeowners insurance is a really big market and when I started Kin, I was really excited because it was $100 billion dollar market. It's only been seven years, it's $150 billion market now.

‍

Bill Cillufo:

Wow.

‍

Sean Harper:

Really unusual that you have a legacy market that has a CAGR that high. And the two things that are driving it are increased investment in housing and increased weather volatility. I think both of those are pretty enduring trends. That's the market we're in and that's a market that's by, and large occupied by an oligopoly of 40-ish insurance companies. There's a long tail, but mainly it's 40. So, it's not a very competitive market and there hasn't been a lot of innovation.

‍

So, the innovation that we're bringing is three parts. The first is a business model innovation, if you could call it that. I think it's pretty simple. We go direct to the customer. 95% of homeowners' insurance is sold through these local agents. There are actually more than 400,000 retail insurance agencies in the U.S. It's more than four times the number of bank branches, and it's more than two times the number of fast food restaurants, believe it or not.

‍

Bill Cillufo:

That's frightening.

‍

Sean Harper:

Isn't that nuts? And people actually go to their bank. Nobody goes to their insurance agent ever. Really, you could camp outside the office and the only person going there is the owner of the agency. So, 95% of it is sold through these agents and that creates some issues. The first is they're very expensive. That branch network costs about 20% of the premiums to maintain for the insurance carriers. It's a huge extra cost. The second is customers actually don't really like it. If you survey customers and say, hey, would you rather buy from an agent or would you rather buy directly? 70% would prefer to buy directly. If you say, are you willing to pay extra to have the local agent? Almost none of the customers are.

‍

Finally, by going direct to the consumer, we can actually use marketing as a method of risk selection. We'll get more into this later, but there's a huge variance in the types of houses and how... In homeowner's insurance, most of the events, most of the bad things that cause a loss, it's weather related. All the houses are going to get hit by the same weather, but they're all going to respond differently because the homes are built in these idiosyncratic ways. Your roof is different than my roof and my bathrooms are different than your bathrooms and all of that. We can actually pick and choose the customers that we know are living in homes that are more resilient to the weather and we can specifically market to those customers. That's the business model innovation. We're direct to consumer. It helps us with risk.

‍

Bill Cillufo:

Whereas the folks that are distributing through agents can't do that because they take whoever the agent goes and finds and the agent is probably not incented to do that. Right?

‍

Sean Harper:

That's right. And they have very crude ways of doing it. They'll tell the agent like, hey, your portfolio looks like X. We'd like less of this and more of this, but it's not very nuanced the way they can do it. They don't actually have control. They just have influence. The second thing we do differently is tech innovation, which plays very well with the business model innovation. The technological innovation is two parts. First, we've built, from scratch, the best, most cutting edge efficient insurance core processing system. It's proprietary to us. Most of our competitors don't control their own core processing systems. They're outsourcing it to a company like Guidewire or Duck Creek, which for the most part are not very good software. It's software that's 30 or 40 years old. For the most part, it's still on-premises software. It's really, really not good, so that gives us a big advantage.

‍

The second part of the tech innovation is all the insurance companies, us included, we all have these smart actuaries pricing [inaudible 00:05:04] and stuff like that. They're crunching the numbers. They're saying, hey, for homes that look like this, we expect the losses to be like that and that's how we're going to come up with our pricing. The issue is the way they get all the input data for that, our legacy guys is the agent. Well, that's silly. That doesn't make any sense. The agent has never been to the house. The agent also isn't an architect. He doesn't know about the shingles on your roof. He's an insurance agent. Third, they actually aren't compensated for the accuracy of data. So, what you'll find is the agents are constantly fudging the data because that's their job. They're trying to help the customer by getting them a lower price.

‍

Now, there are lots of bad things that can happen when you fudge the data and one of them is inaccurate pricing. It's important to note that that's a problem at the point for that risk. Oh, if you price this as if the roof was really good and the roof is actually really bad, you're going to be underpriced for that risk. It's also an even bigger problem because it fundamentally breaks their actuarial models, their pricing models. Because if you think about it, they're feeding all of these models and training them on bog input data. So, you actually don't see the correlations that actually exist.

‍

Bill Cillufo:

It sounds like then you must have a process to collect all this data directly from the source.

‍

Sean Harper:

We do. We're manufacturing our own data, and what we're doing is there's... All you have to do is Google your address and you can see there's just tons of unstructured data out there about your home searched images, there's MLS records, there's tax records, there's building permits, there's all this stuff. It's all unstructured. That's the problem with this. So, what we've been doing for the last seven years is training a set of models to take that unstructured data and package into structured data and then training our pricing and underwriting models on that input data. It's not always perfect. You don't always have perfect information, but it's way, way, way better than the alternative, which is just asking some guy who actually has a bunch of reasons where he'll be inaccurate.

‍

Bill Cillufo:

Right. That allows you to basically use this undercut price on the best customers and then lose the worst customers because you know they're not as good as the industry thinks they are.

‍

Sean Harper:

Right, and it's more accurate pricing. And that ends up being a bigger deal in some places than others. Basically, what you find is in areas where there's more weather volatility, having accurate pricing makes a much bigger difference. So, where I live in northern Illinois, we don't have that much weather volatility. And if you know exactly the type of shingle that's on my roof, it's helpful. It doesn't hurt. But if I lived in Charleston, South Carolina or Tampa Bay-

‍

Bill Cillufo:

You probably only care whether somebody has a flat roof or a slope roof, so the three feet of snow doesn't cause a leak.

‍

Sean Harper:

That's right. The roof shape is an interesting one because there's more nuances to that. No roof is truly flat, and the roofs all have different angles on the slopes. You actually want to take that into consideration. There's a lot of nuance in roof shape, since you bring it up, but it really can make a big difference in some of these places. This is one of the reasons why you see the headlines of insurance Company X is leaving Y geography entirely. Okay, well that's because they actually don't know how to tell the difference between the homes that are going to be profitable for them and the homes that aren't. We're able to do that, and it allows us to serve these customers who need us the most, the folks who live in higher weather volatility areas, which by the way is more and more of the country every year, unfortunately. That's what we do.

‍

Bill Cillufo:

Yeah, that's fantastic. Before we go into the next several layers of Kin, I'd love to spend just a minute or two exploring your background. I know you got into the startup world quite early after you started your professional career. Wonder if you can just talk through how you got into the startup world and what was it about startups that really attracted you?

‍

Sean Harper:

I was a tech guy first. I was one of those kids that grew up programming, and I was just really into making these stupid little apps and games and stuff like that. It's going back to middle school. Then I was like, "Well, I'd actually really like people to use the software." That got me more into the business part of it. And then I realized I was actually really interested in business. I was really interested in economics. I ended up going to school and majoring in computer science and economics, which is a cool combination, and I've really been doing FinTech stuff ever since. When I was in college, I did the same investment banking internships that everyone else does, but my first gig out of college was working at BCG. I was mostly working with financial services companies, mostly retail banking, credit cards, PNC insurance. I was like, "Really interesting." But I looked at myself after two years and I was like, "Gosh, Sean, I always thought you were going to be a tech guy. What are you doing making spreadsheets for banks all day? This is stupid. That's not what you're doing." So, I quit.

‍

Bill Cillufo:

Hey, that is high-tech at some banks, come on.

‍

Sean Harper:

Some of my spreadsheets were legendary. I had a lot of macros in these spreadsheets. I quit and I joined a VC firm, and I did that for a couple of years, really early stage. Did two investments there. And I was like, "Oh, okay. This feels better. This is more like what I was aiming to do." When I was working at BC, one of the things I learned is actually, at least in that case, I had a lot more spare time after working as a VC analyst versus four days a week travel at a management consulting firm. I actually started my first company in parallel with that gig. It was an online retailer, it a e-commerce company.

‍

There was a specific product that we were interested in selling, my co-founder and I in that business. It sounds silly now. We were the largest seller of replacement parts for satellite radio systems, which at the time was this huge consumer product.

‍

Bill Cillufo:

I still have one in all my cars.

‍

Sean Harper:

Yeah, Sirius. Yeah, it's actually a really good product. At the time, I had done a consulting case for a retailer, and I realized how much money was in these spare parts because people are relatively price insensitive. You're like, oh, I have this thing and part of it stopped working, and I can't use this. I can't listen to Howard on my way to work anymore because the power adapter for my Sirius satellite radio stopped working. You didn't care if you were paying 20 bucks or 15 bucks to get the replacement part. That was the business that we built. It was really fun. It was a bootstrap business. We were on the Inc 500. It was cool. That was the beginning. And then I did another business.

‍

It was a payment processing company. Similar, I had had some consulting experiences in that space. I had consumed the payment processing product at my prior company, the one I had started. Oh, I see an opportunity here to do it better. The product that we launched was basically Stripe, but without a $20 million seed round. Our seed round was only 2 million.

‍

Bill Cillufo:

Details, details.

‍

Sean Harper:

When we were out raising our Sirius A, we got an acquisition offer, and we took it. And I spent three years at the acquirer, which was Groupon. I built the square competitor within Groupon. I thought that was really an interesting experience, but ultimately, my future was I was going to do something entrepreneurial again, take a bigger swing. I'd hit a single, I wanted to go and swing again. That was when I started incubating Kin and thinking about what ultimately became Kin. I took a year to explore ideas and do some little independent consulting gigs and do some angel investing and stuff like that, but ultimately, kick the tires on a bunch of things. And this was the opportunity that excited me the most.

‍

Bill Cillufo:

It seems pretty logical how you could go from e-commerce to a Stripe-like type thing. I get that you spent a year looking at different ideas, but homeowners' insurance is pretty out of left field, given what you had done so far. What even got you to start thinking of that idea?

‍

Sean Harper:

Interestingly, I had a consulting case at one of the big insurance companies where I spent a lot of time in insurance agents' offices. It was a go-to-market case for this insurance company. I just saw how wasteful that agent experience was, and I saw how weird it was that the actual insurance company would be... One of the questions they asked us was like, "Well, why do customers come to us?" Oh, that's a weird question for you to be asking me, 24-year-old management consultant. Don't you know? They were like, no, no, actually we don't because the customers go to the agents. First of all, these guys weren't really doing field underwriting as they would say. They didn't really know very much about the risks. They were just going to Zillow and copying some stuff off of Zillow. And that was actually the data that was getting fed into these underwriting algorithms.

‍

You also noticed that those agents were really high margin business. They were actually a much higher margin business than the carriers themselves. At the carriers working for, they were happy if they had a 5% net margin year, that was a good year for them. And then if you crunch the numbers on these agents, which are actually all small, below scale businesses, you're like, whoa, these are like 40% EBITDA margin businesses. If you draw the whole value chain out, the agents are extracting the majority of the value. And why is that?

‍

Bill Cillufo:

A bit of an aside from Kin, but do you know why PE hasn't tried to roll up all these agencies?

‍

Sean Harper:

Oh, they have. Yeah, absolutely.

‍

Bill Cillufo:

Oh, I gotcha.

‍

Sean Harper:

There's a huge amount of roll-up activity in PE, in insurance agencies, especially on the commercial side. There are two private companies. There's Acrisure and Hub. These are both $20 billion market cap private companies, PE roll-ups.

‍

Bill Cillufo:

Gotcha. But because there's so many, most of them are still these little mom and pop.

‍

Sean Harper:

You got it. That's exactly right, especially on the personal line side where I think there's been less consolidation activity. There's still quite a lot going on in personal lines, it's just so big. There's so many of these. There's 400,000 of them. There's a lot. I had some context for the problem, and we actually drew the map of all of financial services, and I was like, I'm looking for something that meets a few criteria. I want something where it's overly intermediated. Some things are really complicated, and you need a lot of intermediaries, but a lot of products, especially consumer financial products, are not that complicated and you don't need the intermediary there. It actually makes it worse.

‍

Bill Cillufo:

If anyone can relate to the problem, it's the QED with our Capital One heritage, which that was a big part of the premise too. Why would you get a credit card in a branch?

‍

Sean Harper:

Why would you get a credit card in a branch? Absolutely. Yeah, very much so. Capital One is a big inspiration to us. It was a company that was on our list of, we could be like the Capital One for homeowner's insurance, that would be really cool. We were also looking for something where there were new data sources that could be used for pricing and underwriting. We looked at everything. There was one other idea that rose to the top. It was also an insurance. It was small business insurance. Ultimately, we decided that this was better because it has a larger, more homogenous market. Small business insurance has many of the same problems that homeowner's insurance has. The reason why we didn't want to go after that is because small business is really a bunch of different niches, and you actually do need a different product and a different go-to market to get the yoga studio versus the coffee shop versus the car dealer. We just loved the fact that homeowner's insurance was so big, it was so homogenous. It met our criteria specifically around new data sources being available, around it being overly intermediated.

‍

Bill Cillufo:

Oh, that makes a lot of sense. When you decided to start the satellite radio parts company, did you think about this insurance idea then? Or more just you had done the case, you weren't really connecting the dots until you went back and systematically looked at it later?

‍

Sean Harper:

Because I always have this entrepreneurial thing, I'm always thinking about things I can do. I had actually thought about, when I was at BCG, doing some of these cases. I thought about doing a roll-up in the space, and then after doing my payments company, I realized how important product-led growth could be in some of these things. You think payments as a commodity and it kind of is, but also there really was a real appetite from the customer to have a more streamlined experience, to not have to answer all the dumb questions, to not have a seven-day underwriting period. There's all these nuts and bolts things about the product that was really inconvenient. That led me to think of more of a product angle for insurance. So, I was like, "Yes, I do want to occupy the part of the value chain that's so inefficient, that's so distributed with 400,000 people doing it, where they have such high margin that they don't really deserve, maybe, for the work that they're doing. But also, I want to do that by having a really unique value proposition to the customer.

‍

Bill Cillufo:

You obviously had a big industry. You were looking for a startup. You had seen the problem firsthand. Were there any major obstacles that you saw as you started to think about this and like, oh, maybe I shouldn't go do this, that you had to get over to take the leap?

‍

Sean Harper:

The two biggest question marks were do customers want this? And can we manufacture it efficiently? So, we had to answer the first question, which was do customers want this? What we did is we actually bought a little insurance agency and we wrapped it in a little bit of UX. It wasn't that expensive. It was a few hundred thousand dollars. We wrapped it in a little bit of UX and we started to run marketing experiments. And we're like, oh, okay. After doing this for six months, the unity economics, while they're not great, but if I increase conversion rate here and I increase conversion rate here and I reduce the churn there, then they will be great. They'll go from being okay to being great. Okay, cool. So, we did that. Then the next thing was, okay, well how do I get control of the whole stack?

‍

Because as an agent, we didn't have any differentiated value proposition. That was one reason why our conversion rate was low. How do I control the whole stack so I can manufacture a product that is differentiated? I come from the payments world, and I'd done some angel investing and lending, and I was like, "Oh, we have to find a rent a charter for insurance." And then we realized, oh, that doesn't actually exist. There is this other thing, it's similar to rent a charter called the fronting carrier. They rent out their insurance licenses. It's not as well established in the ecosystem as the rent a charter, but that was our next step. We're like, this is a very difficult industry to approach. It's no accident that it's basically an oligopoly. It is designed to keep people out.

‍

We know we're going to need to climb up the stack one rung at a time. So, okay, next step is let's set up something where we're an MGA, which is like a virtual insurance company. It gave us all the control over the product that we wanted. There were some downsides to being an MGA. It actually didn't give us all the control over the product we wanted. It created a real existential risk for the business because you had this one counterparty that was so important to you. So, we knew there were some downsides, but we were like, well, we don't have the money to start an insurance company.

‍

Bill Cillufo:

Are you an insurance company now or are you still in the process of working up that regulatory stack?

‍

Sean Harper:

We are. I'll talk to you about that in a second. That was the next step. Basically, the first year we were running these marketing experiments. The second year we were like, okay, cool. Now we have this business relationship as an MGA where we can run everything through our software, but we have to hustle to write the core processing system, which is not an easy task to write a good core processing system for insurance. Now, we've rewritten that thing five times since because it was very much like, hey, here's the MVP. We need to be able to issue a policy. I need to be able to take the customer in, get the underwriting info, manufacture the data, print a PDF, basically. Obviously, now six years later, the system does a lot of stuff in addition to that and does it really, really well, but that was the next step.

‍

Then two years later we were like, okay, well we're on the verge of being a scale... This is 2019. Now, we're on the verge of being a scaled business. We have unit economics that are working. We're really worried about this single threaded relationship that creates an existential risk for the business. The MGA relationship isn't giving us all the degrees of freedom that we want. Let's go for the big time, let's start our own insurance company. We now manage two insurance companies. The first one we started was in 2019, so it took a year to get regulatory approval for that. And I had to go raise a bunch of more money because we need to capitalize this thing. You use reinsurance and stuff to reduce the amount of capital that you need, but we still needed 30 million bucks sitting in a bank account.

‍

At the time, we'd only raised a series A, so we only had 12 million bucks. Then we went and spent a lot of that on developers and stuff. So, we had to raise another round. We had to wait a year for regulatory approval for this. In the meantime, our relationship with the MGA sponsor got worse because they were like, "Wait, you're leaving? What the heck? Why should we care about you now that you're leaving?" It's like, "Well, we gave you warrants. You should care about it." It got a little bit tense there, but we eventually did get in. It's interesting, we actually have a novel financial structure where Kin actually isn't an insurance company. What we did is we set up these things called reciprocal exchanges. Those are insurance companies. They're fully licensed, they have capital, they have credit ratings. They're regulated just like any other insurance company, but they're actually owned by our policy holders. We manage them in exchange for a fee.

‍

Our business actually looks like an MGA business where we're just getting this recurring fee to manage it, or it looks like an asset manager almost. You could think about this, we're the GP and our customers are the LP in these insurance exchanges. It's a good structure for us because it gives us a stable, recurring revenue stream. Our profit center is what our legacy competitors cost centers are, the agents and the outside software. We're basically keeping that margin. If you think about what our legacy competitors, their profit center is underwriting income and investment income, and we're actually leaving that in these customer-owned exchanges. So, the customers are literally getting a better deal. And I think this ultimately makes it very difficult for our competitors to compete with us because what's historically been their profit center, I'm just giving back to the customer.

‍

Bill Cillufo:

It's funny, my wife's whole family is ex-military, and so I know she's always tickled every year when we get our USAA rebate check. Probably the similar concept, right?

‍

Sean Harper:

Very similar. USAA is the second-largest reciprocal exchange in the U.S. The largest is actually Farmers. There's another one called Erie that's pretty big. They're in the Mid-Atlantic. The management company of Erie is publicly traded, so it's an interesting company for us to go and look at. But yeah, it's very similar to what you just described with USAA. We don't give the dividend check, we just leave it in there and give the customer a lower price, but you can do either.

‍

Bill Cillufo:

Let me dive into this. I know back when you were giving the intro to Kin, you talked about you're innovating across a number of vectors. I know innovation and insurance has been notoriously slow and I'm sure there's tons of factors behind that. One of which being, correct me if I'm wrong, state regulators have to approve all of your pricing schemes and you've got to show a bunch of data behind it. I'm guessing state insurance regulators aren't really at the cutting edge of large language models and all of the tools that you use to convert your unstructured data to structured data, et cetera. How have you thought about that and how have you overcome that obstacle? And how big of an obstacle is it from say a perfect world where you could go do whatever testing you wanted to do and do whatever you wanted?

‍

Sean Harper:

We found it to be manageable. It is true that in the majority of situations, in the majority of lines of insurance, you do have to have your rates and forms approved by the regulator. An example of that be do I charge X factor if your roof is shaped like a hip shaped roof where all the sides are angled and I charge Y factor, if your roof is only angled on two sides? The roof that's angled on all sides is better because the wind just flows over it better. That's what they regulate. What they don't really regulate is how do I know that the roof is hip or gable? That's where a lot of our big data innovation has come is on, hey, let's make sure we have really accurate input data. Now, dirty secret, if the regulators did regulate that input data, they would find that the majority of the input data being used by the legacy industry is just wrong.

‍

It's literally default values in many cases because the agent is lazy and they just click through and they just make assumptions because there's not always a good way for them to know the exact angles of every part of the roof, just to use an example. We found that to be a really good nexus for innovation is on the input data. The other thing that is nice is as a direct to consumer business, the state regulators don't pay as much attention to who are you marketing to. I can have an interesting situation where I am really good at knowing if a customer has... This is just an example, I'm really good at knowing if a customer has a metal roof. Metal roofs are great, by the way. If you can afford a metal roof, you should get a metal roof. They're great. Then I can offer that customer an actuarially justified, regulatorily approved lower rate because they have a metal roof that's really resilient to hail and storms and everything like that. Then I can create a marketing campaign that specifically markets to the customers with the metal roofs.

‍

We have this closed loop system. We also have customers where we're like, oh, I can tell from aerial imagery this customer doesn't have a very good roof. The shingles are curling at the edges. There's striping on them. We have image recognition algorithms that can tell us stuff like that. All right, well, we're not going to market to that customer. All of our competitors are in the bad situation of, they obviously also don't want those roofs. They don't have a good way of knowing because a lot of times they don't have the same image recognition capabilities that we have. They don't really know which roofs are in one camp versus the other. But even worse, they're like a hockey goalie because the agents are just flinging all the houses at them, and they have to be defensive. And if they let one through by mistake or because they didn't validate the data or because the agent was good at finagling their system, well now they're in trouble. Now they have something in their portfolio that doesn't really belong in their portfolio. The marketing lever has actually turned out to be really important to us.

‍

Bill Cillufo:

In terms of this data advantage and analytics advantage, it's super innovative what you guys have done. Are all of the advances in LLMs and open AI and all of that type of thing, is that helpful for you because it allows you to advance your capabilities, or does it hurt because it lowers the bar dramatically for other people to come in in ways that you really pioneered? Or is it just a red herring and doesn't really matter a whole lot to what you're doing?

‍

Sean Harper:

LLM specifically are a little bit more of a red herring. Now, you do get a surprisingly decent answer. I'll just use an example. If you upload a picture of roof to ChatGPT and you ask it, hey, describe the roof for me. It's a 50% answer, which is surprising because you know that it wasn't trained on images of roofs, or you could upload a building permit. Building permits are really messy data because they're different literally in every town. You could upload that. Similarly, it'll give you a 50% answer, but we really need more like a 95% answer. So, what we've been doing is training specific models on specific input data. LLMS are also really expensive to run right now. So, the specific models, because they're trained on that, have a much higher degree of accuracy. And for us that's important. We can't have 50% accuracy and they're a lot cheaper to run.

‍

Generally speaking, the fact that the tools have all gotten a lot better is great for us. And I think for anyone who's trying to apply it in this way. The APIs that are available right now, I think, are too general for this and we've really been training our own models.

‍

Bill Cillufo:

Let me switch gears to another angle. I know that a big part of your motivation is using technology to really help consumers, and you're in a pretty unique spot to do that. I'd love your take on some of the places you're doing business, Florida, Texas, hurricane zones, et cetera. Customers obviously have the risk every day of a pretty severe incident. But they also have issues, hard to get insurance, how to get insured at a reasonable price, get dropped all the time, et cetera. And you've found an interesting way. I wonder if you can talk about the customer angle of what you guys are doing and how that's really pretty unique in this industry.

‍

Sean Harper:

It was in the early days where... Now, I knew in the back of my head, it's really hard to win in financial services if you go up the gut of the customers that everybody knows are good. It's just hard to dislodge them. If the legacy guys know that this is a really good customer, it's one they all want, you have a hard time attracting that customer, especially for sticky products. So, we knew we needed to find an underserved niche where it was maybe not as well understood by the legacy competitors or where we had a unique advantage. So, we were looking for that. And then the early days of marketing, we really just realized, you talk to somebody about this in Wisconsin where I grew up, and they'd be like, oh yeah, that's neat.

‍

You talk to somebody about this in Florida or Texas or California and they leap for it. They go, oh, that's really interesting. Actually, my homeowner's insurance, it just doubled in price. Or I only have three choices of insurance companies or it's really hard. I have to work with these weird providers, and they have all this paperwork. It's really annoying. So, there was a real engagement there. And we're like, "Oh, that's interesting. Maybe there is a regional element to this." And then we looked at it more and we realized that the things that we were trying to do... You do have a difference, there is less availability of insurance, and it is more expensive in areas where there's more extreme weather. More and more places every day are being exposed to extreme weather, and that's driven by global warming. So, we're like, okay, that sounds like an interesting trend to be part of helping these customers who are underserved.

‍

There's more and more of them every day, that market is growing. And then we thought about it some more and we're like, oh, actually this really maps very well to the things that we're good at. Why? Okay, well, in the places where the insurance is more expensive, the agents, the distributors still get paid the same percentage fee. So, they're making a ton of money. Homeowners insurance companies in Florida might not be doing that well. Homeowners' agents in Florida are doing incredibly well because they do the same amount of work as an agent in Chicago, and they get paid four times as much. Okay, well that's interesting because we occupy that part of the value chain. I want to go there where the economics are best for that part of the value chain.

‍

And then we looked at the stuff we were doing on the data side and we're like, well, if I know exactly the type of shingle that's on this roof that's going to make $100 price difference in Chicago, it's going to make a $1,000 price difference in New Orleans. Obviously, this skill that we have of understanding the data and the physical traits is way more useful in New Orleans than it is in Chicago. That was really the triangulation of data that led us into this. We have a pretty coastal footprint. We're in South Carolina, Mississippi, Alabama, Louisiana, Florida, Texas and Arizona. Arizona is the outlier there. And then the next few states that we're launching, we're going to be launching another batch of states this year in 2024. They're all coastal states.

‍

Bill Cillufo:

When you guys hit Massachusetts, I'll be very excited to try to be a customer with our vacation place on the Cape.

‍

Sean Harper:

Yeah, no, getting insurance on the Cape is not easy, right?

‍

Bill Cillufo:

Totally.

‍

Sean Harper:

There's a lot of places like that. Coastal Massachusetts, coastal Rhode Island, coastal New Jersey, they're all pretty hard places to get insurance. It's pretty expensive.

‍

Bill Cillufo:

We definitely work with the agent, and she has a new company for us every year, and so I'm very familiar with the problem you're talking about. Look, we only have about 10 minutes left. We've skipped half our script here just because the business model of what you're doing is so interesting. But I'd love to ask just a couple questions about the inside of the company and management as we close. You've now grown to roughly 500 companies, it's your third startup. How do you think about corporate culture as you add a couple startups that wound up selling pretty early? You've now probably learned a ton from that, building your first startup when you were small. I'm sure you've learned a ton more as this one has grown and gotten to be large. How do you think about company culture and what are some of the lessons that a listener could take away.

‍

Sean Harper:

Culturally, what's always been really important to us at Kin is to make sure that people who are closest to the data are making the decisions. I just feel like a lot of the times you have a lot of waste, especially in bigger companies, people are like, "Oh, let me ask my boss about that." Okay, well, your boss is probably not as knowledgeable about this problem as you are. That's always been a really key value for us. Another one is we've always had this idea of running through walls, and this is certainly true, to do this is really hard. We do accomplish the impossible every day. And then we had to amend that. It was an interesting one. As we got bigger, we had to change that stated core value to be run through walls together because you'd have these guys who would just be like, I'm going to solve this problem. And they'd go and they'd do it, and it would have all of these unintended consequences.

‍

You're like, well, you did solve that problem, but let's talk about these other two problems you caused, just because the problems we're solving are more complicated now that we're bigger. We've had to evolve over time. We try to be a company that's really low ego and really focused on the numbers. That can be hard as an entrepreneur, especially somebody who is pursuing an idea that is at least somewhat visionary because on the one hand you have to have the vision, but then also, well, don't focus on that too much. Here's what we have to do tomorrow. Do this one small thing tomorrow and then we'll do another small thing tomorrow and then we'll do another small thing the day after. And that's what adds up here. I feel like certainly in our space, InsureTech was the space with a lot of big ideas, but a lot of companies that did stupid things like trying to scale with unit economics that weren't positive. Okay, let's not do that. Let's try to be somewhat practical in our approach here. That's always been a big part of our DNA is just being practical.

‍

Bill Cillufo:

It makes sense. It's a great example of the lesson of the run through walls where one thing works when you're 10 or 20 people, and the core kernel still is important when you get bigger but needing to modify it. Your be close to the data, really resonates really to Capital One. I think we also had a very similar approach. I think I was managing $100 million marketing budget at age 23 or something, which I look back and I'm like, wow, they were stupid to let me do that. But similar concept, the people doing the analytics making decisions. And I think there's also a good example when the company scales creates a challenge because bottoms up companies can just turn into a bit of a gummed up mess to make decisions as you get bigger. So, how to figure that out will be an important one, I think.

‍

Here's another angle. I know when COVID hit you and everybody else went remote, a big debate now of some companies have gone back to almost all in-person, a bunch of hybrids, a bunch of people love remote. You're more on the remote side if I understand correctly. How have you thought about that and is it just a no-brainer good for you guys or are there trade-offs that you wish you had more in-person? How have you thought about that trade-off?

‍

Sean Harper:

There are definitely trade-offs, and we were very much an office company before this. We had an office in Tampa area, and we had an office in Chicago, and everyone was in one of the two offices every day, and we loved it. We had so much fun together. We'd play ping-pong and all that. I would cook people breakfast. It was a huge part of our culture. We didn't have a choice, we had to go home. Just because of how we grew, our company tripled in size over that time period. So, by the time we were able to gather again, we had people all over the country and we're like, well, this is really great on the one hand, because Kin... Especially because we're not in Silicon Valley, and neither Chicago nor Tampa are big talent centers. For us, it was a net positive trade to be able to recruit everywhere.

‍

I think maybe if you're in Silicon Valley, that's not as true because they have such a concentration of talent there. But for us it was really great to be able to open a recruiting aperture. The COVID time was hard on the cultural side because we could never see each other. Now we just get together a lot and we get together on the individual level. We get together on the group level, we get together on the company level. We have offices where it's more like it's a small office, but there's space we can flex into. That's been really nice because we can get a larger group of people together. We're committed to it. I think there's another version of the world where we never went remote because there are costs to it. It is harder to coordinate, it's harder to build culture, but there's no putting the genie back in the bottle for us. We just have too many people in too many different places to say, hey, go back to the office. Half of the company doesn't live anywhere near our offices.

‍

Bill Cillufo:

I spend a fair amount of time internationally, and it's always a bit of a debate of now that you can hire remote people, should you go with a global talent base? Or what are the benefits of having everyone together? And everyone comes up with different conclusions. It's fascinating how broad it is. It makes sense, over time things would shuffle into multiple options and employees over time can self-select into the environment that works. It seems pretty clear you can be successful. As we close, and I really appreciate the time, this probably could easily go another hour here. One thing I see sailboats right behind you on the piece of art. I know you've got a huge hobby for fixing old sailboats. How did that come to be?

‍

Sean Harper:

I don't know. That's a good question. I grew up sailing and then I didn't do it for a long time, and then in COVID I picked it back up again because it was one of the things you could do. I got one sailboat and I sailed that. First, I was just like, "I want to sail. I'm going to go down to the beach and rent a sailboat." But of course, that wasn't open because COVID was [inaudible 00:42:30]. Okay, now I have to buy my own sailboat. And then I just got into it, and I bought another one. I don't know, it's a fun thing to do for me. I like being out in nature.

‍

Bill Cillufo:

That's awesome. Do you ever do any of the long haul Mackinac races? I had a good buddy of mine in high school that does all these Port Huron to Mac, Chicago to Mac, but sailing on Lake Michigan must be pretty fantastic.

‍

Sean Harper:

Yeah, Lake Michigan's pretty nice sailing and it's very close. In Chicago, you're right here.

‍

Bill Cillufo:

I really appreciate the time today. We try to close with the same question for everyone. Hopefully we have a number of aspiring entrepreneurs listening to these. Love to know if you were to give one tip to an aspiring entrepreneur given on your third rodeo right now, what would that be?

‍

Sean Harper:

Figure out the unit economics first. It's the thing. So much time is wasted, so much money is wasted scaling businesses that don't have positive unit economics and never will. We only have so many years on this planet. Let's find something that's sustainable and then figure out how to scale it.

‍

Bill Cillufo:

No, that's awesome. That was a lesson that the world forgot in 2020 and 2021, but I think it's quickly trying to remember again. Very much agree with that. Sean, thanks so much for joining today. It's been a great conversation. We're huge fans of Kin and love the innovation that you're doing here, so really appreciate you spending the time.

‍

Sean Harper:

Thanks for having me.

‍

Bill Cillufo:

And to all you listeners, till next time, take care and thanks for listening.

‍

This has been the FinTech Thought Leaders podcast, your window into the world of venture capital and financial services with today's digital disruptors. QED is proud. To provide the best FinTech advice you can get. To learn more or to read the full show notes from today's episode, check out qedinvestors.com and be sure to also follow QED on Twitter and LinkedIn at QED Investors. Thanks for listening.

‍