What It Takes To Build a Trusted Brand and Profitable Insurtech Unicorn

Headlines love the drama of insurers pulling back, but the real story is smarter underwriting meeting a changing climate. 

In this episode, Julie Verhage-Greenberg from This Week in Fintech, and Emmalyn Shaw,  sit down with Sean Harper, Co-Founder and CEO of Kin. to explore how a data-first, direct model turned a tough category into a profitable growth engine—and what it takes to insure homes when weather risk rises a bit every year. The conversation starts with the broken parts of homeowners insurance: 400,000 local brokers guessing about roof age and materials, misaligned incentives that reward underreporting risk, and claims processes that drag on when losses are large. Then we dig into how Kin rebuilt the stack using MLS records, digitized permits, aerial and street-level imagery, and machine learning to extract real home traits and price risk precisely, especially in catastrophe-exposed regions.

We compare growth-at-all-costs to measured scaling with unit economics that actually hold up years later. You’ll hear how direct distribution lets Kin manage portfolio spread—steering demand away from overexposed neighborhoods—and why that matters when entire blocks can burn or flood. We also highlight the customer experience: clear coverage trade-offs, faster claims, and communication that blends automation with human judgment when empathy counts. Generative AI now powers back-office efficiency, drafting compliant letters and tightening timelines so G&A barely moves while revenue grows.

We don’t stop at the model. We examine the 2022 reinsurance squeeze driven by inflation and rising rates, how pricing reset across the market, and why reinsurers increasingly reward accurate underwriting over blunt cat models. Looking ahead, we share priorities: deeper cost compression, expanding into homeowner-adjacent products like mortgages, auto insurance, and home equity, and financing safety upgrades that cut both premiums and losses. If you care about the future of insurtech, climate risk, and building products customers actually trust, this one delivers both playbook and perspective. Enjoyed it? Follow the show, share with a friend, and leave a review to help more listeners find us.

Flourish Ventures is an $850M global early-stage venture firm that backs entrepreneurs transforming financial systems for the better. Its portfolio spans more than 100 companies across the U.S. and emerging markets. The firm also supports innovators shaping policy, media, and research to accelerate lasting change in financial services.

This Week in Fintech (TWIF) is the largest fintech community in the world, presenting news, podcasts and newsletters from around the world.  Subscribe to our newsletter to stay up to date on the latest in news opinions, and all things financial technology.

Transcript

Setting The Stage: Why Insurance

Julie Verhage-Greenberg 0:03

Hey guys, I am Julie Verhage- Greenberg with This Week in FinTech, and welcome back to our special podcast with Flourish Ventures. In this special series, we are spotlighting the people and ideas reshaping the US financial system.

Emmalyn Shaw 0:17

And I'm Emmalyn Shaw, co-founder and managing partner at Flourish Ventures. And today we're diving into one of the most challenging categories in fintech, insurance. And we're going to talk about how Kin became a profitable unicorn at a time when the insurtech sector has been under real pressure.

Julie Verhage-Greenberg 0:34

Sean, as we were mentioning before the show started, I feel like we know each other because from my work with Flourish and Commerce and BTV and others, like they all talk about you and how amazing Kin is doing. So I would love for those that are listening that maybe don't have those interactions, for you to just give a quick background on yourself and Kin.

Sean Harper 0:55

Cool. Yeah, thanks. We have a lot of good friends.

Julie Verhage-Greenberg 1:00

You've done a good job picking your investors.

Homeowners Insurance Becomes A Hot Topic

Sean Harper 1:03

Yes. Well, we're we we love our investors. Um we started Kin nine years ago, almost eight and a half years ago. And we use technology to provide homeowners insurance. And homeowners insurance is super important in people's day-to-day lives. You if you own a house, you pretty much have to have it. Um if you don't have a mortgage, theoretically you could go without it, but it's not wise. Uh, if you have a mortgage, the bank is gonna make you have it. Um, so it's it's a really big part of people's day-to-day lives. And when we started the company, it wasn't like a super interesting area of the economy yet. Like people weren't paying attention to it. But nowadays it it very much is.

Julie Verhage-Greenberg 1:48

Like, it's almost every week there's some newspaper article in one of the big national papers about uh homeowners insurance and how I was gonna bring that up because I was gonna say I feel like over the last decade there's been just like disaster after disaster after disaster.

Sean Harper 2:04

It's become a hot topic, and it really is at the intersection of a lot of the most important trends that are happening, right? All the crazy stuff that's been happening in the housing market, inflation, global warming, weather, volatility. Um, and so and and and as in part because of those trends, the market is growing very quickly. So I come off, I've been doing FinTech for a really long time. And the um before this, I had a payments processing company that I sold. Super small time exit. I was eager to do it again. And one of the things that attracted me to homeowners insurance was just how big it was. It was a $100 billion industry. Little did I know that it would double in size over the next eight years. It's now a $200 billion industry. Um it's a really important problem. And it's an area of the economy where there are a couple of big shifts, right? There's there's a change in consumer preferences. The way this product is sold is through 400,000, I'm not kidding you, local insurance branches, physical insurance stores to sell this virtual product.

Too Many Middlemen, Bad Input Data

Julie Verhage-Greenberg 3:14

I wanted a question for you too, because like I I own two homes now, technically. The housing market in Austin is not very favorable to sellers right now. So we're renting the one what when we bought the other one. But like we we went with the insurance provider that the mortgage company and builder recommended just because like they gave us a sick deal, right? So how does Kin sort of like interact in that part of it since you guys don't also have like your own home builder that's built? And like I went with Taylor Morrison home funding and their insurance provider and stuff, you know?

Why Sean Chose Insurance After Payments

Sean Harper 3:47

Well, that's it's somewhat uncommon. Um now, yeah. So when when you do buy a home, the builders they often monetize insurance, so they have their own insurance brokerage internally, and being an insurance brokerage is an incredible business, so it's like a nice ancillary revenue stream for them. Of course, the majority of homes aren't being sold by a builder, right? They're being sold by one person to another. And you have a lot of choices, right? And you could go with an insurance broker, uh sort of traditional one, one of these 400,000 local stores in every strip mall in the United States, pretty much. Um, there's there's actually only 90,000 bank branches in the US. So that's weird. Like, why do we need 400,000 insurance branches and only 90,000 bank branches? There are only 200,000 fast food restaurants in the US. Well, that's weird. Like, are people really going to insurance stores twice as often as they're going to eat fast food? Probably not. Um so it's a really antiquated way of selling it, and it's very expensive and old-fashioned, right? So the the those insurance brokers, they actually collect a 15% or higher ongoing commission. And so a lot of people will be like, Yell, I like my insurance broker. He's like my cousin's brother's friend, or he's like the guy on the softball team. It's like, well, that's cool. Like, do you like him enough to pay him 15% of your home insurance cost forever? If so, cool, you know, go for it. Um so that's that's one issue, is just the product is distributed in a way that customers, like most customers, what they really want is just to be able to buy it on an app or on a website or over the phone or whatever. They want a digital experience and they're really not getting it. So that's that's sort of problem number one. Consumer preferences have shifted away from the way the product is being sold. Problem number two is and this this is um less of an issue in your case where they're buying it from a builder, because the builder doesn't know about the home. They know a lot about the home, they built the home. But the way the industry operates is not only is the insurance agent responsible for selling the insurance, they're actually responsible for gathering the underwriting data about the house. And in the case where they're not also the builder of the home, they really don't know. And this is a huge problem. It creates a garbage in, garbage out problem because all these insurance companies are all doing really complic, like it's actually not that complicated, but they're doing really good math, right? They have actuaries, and the actuaries are smart, and they're basically figuring out here are all the things we knew about the risk about the home, and here's what happened to it. And it's just a correlation analysis between those two things. The problem isn't the math, the problem is it's garbage in, garbage out. Because they're asking this dude in the strip mall what kind of shingles are on the roof, what kind of bikes are in the house. Uh you know, he he doesn't know, right? Um, and he has an incentive, a financial incentive that's not aligned with that of the insurance company. Right? His his duty is actually to the customer to get them the lowest price. Um, that's their value proposition. You say, Why should I go with you? Well, I'll get you the lowest price. Um part of the way they do that is actually by lying about the home and making it seem safe.

Julie Verhage-Greenberg 7:12

Is that legal or is it like quote unquote legal?

Sean Harper 7:15

I mean, they're they're making their best guess, right? Um, you know, now now technically speaking, it's actually the insurance problem, right? The insurance agent is gonna fill out pages and pages and pages of information about this house. It's gonna get packaged up into an insurance application that then gets sent to the customer and they're gonna sign it. Okay, do you actually read that application? Do you actually know? Like, no, of course not. Um, and so it can end up in a situation where there's coverage dispute because the home, the the insurance company is like, look, your you said your hope your age your roof was four years old. It's not, it's 15 years old.

Julie Verhage-Greenberg 7:58

Um best case scenario, I have Chat GBT read that document for me and like tell me if there's anything I need to look over it.

Sean Harper 8:06

Yeah, um, and so but that sort of coverage uh dispute is uncommon because the there are a lot of regulations. The insurance companies obviously they don't want to get in a dispute with their customers. So actually they just pay the claim.

Julie Verhage-Greenberg 8:18

So what gave you what gave you the idea to start Ken then? Because like going from payments processing to insurance, which like no matter whether it's like home or health or whatever, like insurance just sounds like one of the craziest financial industries that you could possibly get into.

Sean Harper 8:38

Well, it's not that different.

Julie Verhage-Greenberg 8:39

Um from payment processing, you mean?

Sean Harper 8:42

Yeah, it's there there are a lot of similarities. You know, one of the things that attracted me to payments was the there were too many middlemen getting paid too much. Right. This is what almost it's like 16 years ago now. That's a while ago. Um, you really didn't have product product-led growth. Right. So the way you won payments was by having better salespeople. Right? There are these ISOs in in independent sales organizations that were very similar to the insurance brokers, and they were the guys that were making most of the money. The product was very not differentiated, right? You get global payments, you get five surf, like you get first data, whatever. Um, there's no product differentiation. Um, and so I was like, cool, we could like actually cut out these middlemen, make a product that's superior. The product we had was somewhat similar to what what Stripe is now. Um say, cool, we're actually just gonna go direct to the customer with a much better product, and that'll that'll be a better outcome for everybody. So coming out of that, I was like, hey, let's just look at every financial product and see where that opportunity exists. Right? I was looking for too many middlemen getting paid too much, and I was looking for this underwriting input because that was a big part of what my payments business did too, is we were actually better at underwriting the risk because we knew more about the risk. And we were we knew more about the risk because we had a machine. Now, what we could do back then is not at all as cool as what we can do now, but we were able to underwrite the merchant because we were able to figure things out about the merchant using technology that you couldn't by just asking this third-party salesperson, right? You had the same issue. The third party salesperson actually wanted to make the business seem safer so they could get it under it.

Julie Verhage-Greenberg 10:26

And you're saying that all translates to to insurance and what King can do.

Full-Stack Strategy And Differentiation

Sean Harper 10:29

Yeah, it's very similar thing in homeowners insurance. So I I basically just started with a map of every financial product, and we looked at like what fits this uh pattern the best. Too many middlemen, inaccurate input data, no product differentiation. And we looked at every financial product. I gave myself a year to do it basically. And homeowners' insurance met that criteria really well. We talked about there's too many middlemen, that's not how customers want to buy. Uh, middlemen make a ton of money. The input data is bad because you're trusting the sort of unreliable reporter, which is the insurance agent, to tell you about it. So cool, we could teach this as eight years ago. Obviously, we've come a long way since then. Hey, we could teach a machine to go out and find information about homes, and that would be an unbiased and much more accurate source of information. We could train our underwriting and pricing algorithms on that. And we could really differentiate around the product experience. We can do things like we'll actually notice when the weather is bad by your house, we'll text you to ask if you're okay, and you can file a claim via text message. So you address, especially when there are a lot of claims all at once, we've dramatically decreased the cycle time and continue to decrease it around identifying which customers have damage and then ultimately getting their home repaired. So it was they're not that different, a problem, actually. And that's one of the cool things about for me. I've always really loved financial technology because you know, I grew up programming, I'm sort of a nerdy guy. Like, what are the most important bits and bytes there are? Well, it's money, right?

Julie Verhage-Greenberg 12:11

Um it's and I want I wanna I want to turn it to you for a second because you invested in Ken fairly early on. Did his idea make sense to you right away and sort of also align with this theme that we're talking about in the series of like how financial infrastructure is broken in many areas of our ecosystem?

Emmalyn Shaw 12:28

100%. I mean, I think what was interesting from Ken's perspective is obviously they were at the forefront of leveraging kind of distribution, traditional distribution. But what they immediately did that was beyond just kind of you know using digital means to access and acquire customers was they actually knew that there were some markets that particularly were impossible, very difficult to underwrite. So catastrophic regions where there, you know, high degree of variability as it related to weather, et cetera. And they leveraged, as Sean just mentioned, the technology that was available to really underwrite that risk in a way that hadn't been done and to be able to provide a much better, broader service. And I think that approach transformed how people interact today with insurance. I think, and I think that was a it, they were such early, they were so early in identifying the proof points in the technology that could actually help provide transparency, better pricing, et cetera, and then streamlined the entire customer service experience. And I think for us early on, that was the vision. Obviously, it was we were part of the seed, so that was very much the vision. And obviously, Sean, the proof speaks for itself and the performance, but how he has evolved into that, I think, has been particularly powerful. One question I wanted to ask Sean, because I do think you know, a lot of people have tried to play the insurtecn game over the same duration and have honestly been met with a ton of um friction and quite frankly some failures along the way. And and Ken, you know, there's so many decisions, I think, that really that were very thoughtful, but maybe one that's worth pointing out, and Sean, I think for the benefit of the audience would really appreciate it, is you know, you guys focus, you focused on distribution. A lot of the companies were distribution and like growth at all costs. And you were pretty clear around like, hey, we're gonna be measured, and actually we're so well positioned because of our technology advanced, we're gonna underwrite our risk directly and take the reciprocal and you know, really get the benefits of being your own underwriter. What convinced you that the full stack was the right way to go? How how did you make that decision, maybe for the group? And and and what were some of the trade-offs you considered at that time?

Profitability, Unit Economics, And ZIRP Lessons

Sean Harper 14:28

Yeah. Well, so it's it's a funny conundrum because the part of the value chain that is a business person you want to access and participate in, the highest margin part of the value chain in insurance, where distribution is completely decoupled from risk, is the distribution part, right? The distribution businesses are exceptionally high margin. Um, right, like and they're very pedestrian, they're very like boring companies that have been around for a long time. A company called Brown and Brown, their ticker is BRO, super sleepy company. They have like 40% profit margins. They don't do anything all that special. It's like I'm not throwing shade at them, it's an incredible business. Um so you want to do, you want to capture the distribution economics, but in order to have some, but then then you're like, okay, well, like I need to have something different to sell these people, right? So I think a number of people they're like these various comparison shopping things, right? You have companies like you know, maybe like Policy Genius or um Zebra is one, right? There's like more comparison shopping. And the problem with that is that you don't actually have any differentiation. The customer is like, okay, cool, you gave me a bunch of options. And actually, that was something I learned in my last business because we started out as sort of a comparison shopping business and then pivoted into manufacturing our own product, right? We wanted to actually have a differentiated product experience. If you're just comparing other people's options or just sort of brokering other existing options, you really aren't solving the whole problem for the customer. What they want is they want something completely new that breaks all the compromises. And it's especially true because there are a lot of compromises made in insurance, like because the distribution is so decoupled from product from manufacturing, um, there's a lot of inefficiency there. And actually, if you hang out in a traditional insurance agency's office, which I did many years ago, like when I was a consultant, my first job out of college, I was a consultant at BCG. And for whatever reason, I ended up spending a ton of time on this case where I was like actually hanging out in insurance agents' offices and like figuring out what they did all day. And one of the things I discovered was that they actually spent very little time talking to the customer. They spent a lot of time navigating all of the bad IT that the insurance companies they represented had. So you have funny things where it's like, hey, I represent 10 insurance companies, but this was a while ago. This insurance company, their website only works on Internet Explorer, and this one only works on Firefox. I have to have like two computers on my desk. Like, well, that's crazy. Uh right, or like, um, or like where the insurance companies would at would want them to answer a lot of questions. That's like, well, you could just hit like Zillow's API and you can save them all the time, right? Um, and so uh we really wanted to have product differentiation. So it's like, okay, we need to be able to have the infrastructure and manufacture a bespoke product that's for our customers. And that includes the data that's being used for underwriting, includes the claims experience, includes servicing experience, includes the acquisition experience. Um, so so that was why it was important for us to control the whole stack. Uh we don't we don't we never wanted to be a company that was like solving only part of the problem for the customer. We solved the whole problem for the customer.

Julie Verhage-Greenberg 17:51

Yeah, that makes a lot of sense. Um Kit also became one of the only insurtechs e that you, you know, you reach the unicorn status while being profitable. So how critical was it to invest in, you know, data, analytics, et cetera, early on? And then what do you think, you know, other entrepreneurs might have misunderstood about the category that you know you did use and it unlocked this new path for you?

Data Moat: From MLS To Imagery

Sean Harper 18:17

Well, I think I think whatever you're this is a very basic answer. Whatever you're doing, you need to make sure you have product market fit. And when you're doing something with risk, it's especially important because there's this other thing where it's like, well, sure, you could have product market fit, but also be like giving the customer too good a deal, right? To the point where your margins are negative. And so I I think you basically ran into two problems. So and and also by the way, like innovation is supposed to be messy, right? Like, there's not like I'll hear it all the time from like people in the insurance industry, and they'll be like, oh, all these insure techs didn't work. Like, there is no innovation insurance. Like, that's silly. Like, that's just the nature of startups, right? Like VM knows this, like any VC knows, like some of them work really well, the majority of them don't. That's how innovation happens, right? And the ones that do work really well, they have an asymmetric impact, right? They're the ones that bring us forward into the future. So it's like you wouldn't expect 10 out of 10 to work, you'd expect maybe three out of 10 to work. And I don't think that insurance is all that different. Um, now you want to make sure that you have something that speaks to the user, right? Where like you actually have product market fit. And then the second is you need to make sure you have positive unit economics. And I think that actually there's like a confounding variable here, which is that I think a lot of us people that were working in fintech or saw what happened in fintech in like the 2016 2017 timeframe, we're like, oh, a lot of the stuff in payments and lending, well, it's gotten very competitive. There aren't there's not that much opportunity, right? Oh, insurance, it looks Kind of similar. Let's go over there, right? And of course, what happens at that time frame is you you end up, you know, a lot of these companies are series B, series C companies, and they're in the ZERP era now. Okay, so now capital is really, really cheap, and really the investors do want you to do growth at all costs. Right. And so I had like a ton of really frustrating conversations because our investors are all we're always, I think, a little bit more conservative. They're like, look, you want to have, and this is my orientation too, like, don't scale something unless you're sure the unity economics are good, right? Like figure out product market fit, figure out unity economics, and then scale. And you had a lot of companies that were scaling that just didn't have positive unit economics. But I think I do I do think the one thing is unique about insurance is the financial equation is somewhat complicated because a lot of your costs come later, right? You can underwrite an insurance policy and realize only two, three, four, five years later that you mispriced it. And so you really need to understand the it is harder to understand the unit economics, I guess.

Julie Verhage-Greenberg 21:08

And so I was gonna say it's not the same as like you know, building a little product that you sell on Etsy where like you know exactly what your profit margin is and all that kind of stuff. Yeah.

Sean Harper 21:17

Yeah, you make a hamburger, you know what the cost is the cost of beef beef is going in, right? Um, which we don't we don't always, right? Now over time you get you collect data and now you do, right? Um, but at the time you had a lot of companies that were scaling the economics that just were negative. And that led to a lot of capital being evaporated, and I think some of the higher profile like unsuccesses in the market could have been avoided, you know. And also, by the way, some of those were good ideas. They just didn't like properly iterate on it in order to reach the right equation before scaling it, right? So you always want to make sure you don't scale some. That was a crazy time.

Emmalyn Shaw 22:00

I remember that. That was a crazy time. Hey, Sean, can I can I ask you? I mean, so Ken now protects over 100 billion in insured property value, and that was up from like 10 billion, what feels like a couple of years ago. So it's been pretty incredible growth. Um, at that scale, can you talk a little bit about the compounding data and underwriting advantages that emerged? Because I do think we've always been data first, and I think that's something that also was really not um people just hadn't invested and hadn't brought that level of like uh technical talent into the to their house as early. And I think it's something that really started us off in a strong footing, but I think fast forward to today, I think puts us in a very different landscape. So I'm just curious if you could talk a little bit about it from your from your perspective.

Julie Verhage-Greenberg 22:45

One other question to add on to that on the data point of it, is the data messy and hard to understand? Sort of like, you know, payments data can be really messy and under hard to understand how it shows up on different credit card statements, or if it's a tip on Uber, or if it's the ride on Uber, et cetera. Is it the same with housing data, or is it a little bit easier to, you know, digest and make use of?

Sean Harper 23:07

The thing that makes the thing that makes this problem fundamentally difficult is that all of the homes are so idiosyncratic.

Julie Verhage-Greenberg 23:15

Right.

Climate Risk And Underwriting Precision

Sean Harper 23:15

So the same, if you have a block of houses, you're gonna have the same weather is gonna impact all the homes. But they're gonna have completely different outcomes because they're made out of different materials by different people at different times. And there's thousands of things that go into making a home resilient to the weather. And and that's what we wanted to figure out is like, hey, could we know the most about the home? And we had this hypothesis that we could have better input data, right? Can we know the most about the home? Because we saw the MLS that had gone online. We saw a lot of government records that have been digitized, right? Property tax records, building permits. And when you saw an explosion in the capability and volume of imagery, right? So we use a lot of imagery. There are a lot of image recognition algorithms. We're ingesting gigabytes and gigabytes of images of the home from different angles, from the air, from the street, and we're using it to discern traits of the home. Right. So you're the issue here is that the data is very unstructured, right? It's an image or it's an unstructured text file, and that doesn't lend itself very well to actuarial analysis, which requires numbers, right? Um, and so so that's that was sort of the basic idea is just could we know more about the house than everybody? And we what we didn't have, right? We didn't know we we had, if you think about an equation, we have the independent variables and the dependent variables. So we were really good at the independent variables, right? Because we had all this like machine learning and everything telling us like, hey, we know exactly what type of shingles are on this roof. We know exactly the slope of every angle of the roof. And the other insurers, they're just like asking some insurance agent, and he doesn't really know. So he's gonna kind of guess. And by the way, he's gonna fudge the data, right? So we we had that big advantage. But what we didn't have was our own claims database.

Julie Verhage-Greenberg 25:11

Right?

Portfolio Balance And Direct Distribution

Sean Harper 25:12

So we had to basically bootstrap the algorithm off of industry level data for the dependent variable. And so we had a very imperfect equation, right? We were really good on the independent variable, the input data, and we're not that good on the dependent variable. We're only as good as the industry, and actually, probably a little bit worse, because when you buy industry aggregated data, it's never at the granularity that you want it. Okay. Well, now we've been doing this for quite a while, and we have our own um output data. And so we're able to actually get a lot more sophisticated on the algorithms themselves, and we've trained these algorithms on a much richer set of input data. Right? So if you ask, like, you ask like our legacy competitors, these guys, our and our average competitor, by the way, has been around for a hundred years. That's not our oldest competitor, that's our average competitor, is more than a hundred years old. You say, hey, like, what about these kin guys? They'll be like, oh, that'll never work. We have a hundred years of data. Right? And what I would say to them is like, well, sure, but actually, what you really have is an algorithm that's trained on a hundred years of faulty input data. That's not a good thing, that's a bad thing. Okay, so we only have 10 years of data, but it's first of all, it's far well above statistical significance. And second of all, it's trained on accurate input data. So we actually are the only people that have an accurate understanding of what traits of a home lead it to be resilient to today's weather. And that's really, really important because the weather is getting worse. And so when you have these legacy insurance companies leaving these markets or scaling down in these markets, which by the way, are most of the housing stock of the United States is within a short distance of the ocean. And by the way, that's where most natural catastrophes happen, is near the ocean. And the reason why is they are fundamentally unable to tell which homes are more risky and which homes are less risky. They've gotten away with that for a hundred years. Because if everybody in a risk business, if everybody is not good at understanding the risk, they'll just pass along the price to the customer and it's fine. Like they'll all make money, right? The only thing.

Julie Verhage-Greenberg 27:28

I have a question. We've talked a lot about like behind the scenes, what happened, like how like you're able to run insurance more profitably, insure things more profitably, make it a better experience, etc. What about for me as like the end user, if I had Kin as my insurance provider? And then again, like kin aside, what do those people do that live in areas where like, I mean, I've heard of some people now, like they literally can't find housing insurance because these natural, whether they be in Florida or California where the wildfires were, anything. Like, what does the future of that look like? And how is Kin, you know, helping drive that forward?

Sean Harper 28:02

Well, the weather is getting 2% worse a year.

Julie Verhage-Greenberg 28:05

Which doesn't sound like a lot, but compounding that's a lot.

Claims Experience And Coverage Clarity

Sean Harper 28:08

It does compound over time. Uh, yeah, but it's also it's not like an emergency, right? Like it's something that humans have to adapt to. Um, but you'll hear all these things, right? The press is always like, on the on the one hand, they're like, there is no global warming. Or on the other hand, they're like, it's catastrophic, we need to move to the moon or whatever. It's like, well, no, actually, just yeah, there's some places we shouldn't probably live, right? Like Barrier Island. Okay, buyer beware if you're moving to a barrier island. Like they've always flooded, you know, like let alone global warming. Like barrier islands have always flooded. Like it's right there in the ocean.

Emmalyn Shaw 28:46

Sean, what what what would you say Kin is able to underwrite that I don't know how we want to quantify it, but where traditionals just fundamentally couldn't by virtue of all the things we just described. Like, is there is there a order of magnitude of what what we can truly just underwrite in a way that they they wouldn't be able to? Or certainly certainly at a price point that would even be tenable. Let's put it.

Sean Harper 29:10

Yeah, I mean, for for us, it's it's really around identifying the homes that while they are maybe a riskier area, they're actually the safer home.

Julie Verhage-Greenberg 29:18

Yeah, but they're like fire retardant and stuff like that, like the materials and everything that they have. Yeah, yeah.

Sean Harper 29:23

Yeah, the materials and also just the microgeography around the home. Um, but but yeah, the materials make a big difference. The way the home is built, a lot of people don't realize that if you're the eaves of your roof, right? If they're not screened in and you're in a fire area, little embers will go up into your attic and they'll start your attic on fire without you even realizing it. Right, there's all this complicated stuff around is pretty well like building science is pretty well understood, right? But but where where the understanding is lacking is in being able to programmatically identify which homes, the traits of each individual home. Right. So it's it's not so much that like, and actually, by the way, I'm in a lot of the in a lot of the areas where it's like the most most risky, there are other insurance companies that are actually providing insurance at too cheap a price.

Julie Verhage-Greenberg 30:14

Right.

Sean Harper 30:15

So you see that happen in some places where like like Pacific Palisades was a good example of this. Like we were open in the Palisades when it happened. But our price was like three times the price of the market. Okay, State Farm had a price that was lower than the market, and guess what? State Farm ended up over-indexing 2x or more into the Palisades, and that was a really big problem for State Farm.

Emmalyn Shaw 30:37

Um, and so you it's and a lot of those guys dropped as a result of that mispricing, dropped them altogether, right? And that was the thing that that's when the pendulum, I think, swung the other way, which was really problematic.

Sean Harper 30:48

That's right. Then they were like, they realized way too late they were over they they were they were oversaturated there. Um, and that's another advantage of the direct-to-consumer model, is in insurance, it's not just about the expected value. Like you can in you could insure the three safest homes on a block. But or say you insure you insure the three safest homes on a block, but the whole block burns down, it's still not a good outcome. So so because we're direct-to-consumer, we can actually specifically target customers using our ad spend and say, like, okay, well, we just we just signed up a customer in Santa Monica. Let's have the algorithms, okay. We're not gonna spend it in the neighborhoods around Santa Monica, we're not gonna spend as much money. Instead, we're gonna spend more money in Jacksonville, Florida. By the way, the same bad thing isn't gonna happen in Jacksonville, Florida as happened in Santa Monica, California, right? So being able to really balance the portfolio out and create adequate spread is really important too. And that's that's an area where the legacy model doesn't work because it's so so um right, they're they're all selling through branches in your neighborhood. So, what happens when you have a good branch? Oh, you end up insuring the whole neighborhood. Okay, what happens if the neighborhood burns down? Like, okay, that's not good. Um, so that's that's another advantage.

Julie Verhage-Greenberg 32:07

Yeah. And then what what I was gonna say, what about the end user experience too? Because I haven't had to do um a claim on housing yet myself. My husband did once, and it's not it was a little bit of a pain in the butt. Um, but when we were renting, we did have to do um a claim to our renter's insurance, which was lemonade, and it was a freaking great experience. Um, I don't know that lemonade actually ended up ever making money on us or anything. That's a different story. But the experience of filing the claim was phenomenal, like so easy, just digital, etc. I assume that it's fairly similar. And Ken, you mentioned like getting the text message to check if you are okay. And if you're not, you can file the claim right on your phone. Um, so walk me through that a little bit.

Generative AI In The Back Office

Sean Harper 32:50

Well, uh the claims we deal with are a lot more complicated than those. Right. Right. The renter's insurance, it's like what can be something got stolen or yeah. It's mostly your stuff because renter's insurance policy doesn't cover the building itself. With homeowners, it's like something really bad happened, right? Like your roof needs replacing or your home got flooded. Yeah, something like really bad. Like the average severity of our claims is like $30,000. It's like something pretty bad happened. Right. Um and it's complicated because it involves the home. But we have made it quite a bit more digital, and our cycle times are quite a bit faster than the rest of the industry.

Julie Verhage-Greenberg 33:33

Yeah, that was the other thing. It did take quite a while too, because we we had a flood in our bathroom, which that wasn't terrible, but it did take a long time, and it's like it needed to be replaced, so it's just annoying that it's taking forever. And then the roof was just a crappy claim because they're like, oh, well, we don't protect against that, so you're only gonna get like a hundred dollars towards your new roof, and it's like a ten thousand dollar roof. I'm like, thank you, you did nothing for me.

Sean Harper 33:55

Well, that's and that's um that shows how the customer experience when you're signing up is pretty linked to the the claims experience. And uh it's quite often the customers they actually maybe, maybe if you were given a really clear choice, you would have chosen to pay more for a policy that would have covered your roof more in a situation.

Julie Verhage-Greenberg 34:18

Right. Like I feel like we don't ever get like options necessarily. Like with health and care health insurance, you get like, okay, pick like the most expensive plan, but you're gonna get more coverage on different things, or pick the least expensive, but if something goes wrong, you're gonna have to pay some more. I don't remember getting those options with home insurance. And she's like, okay, here you go. Take it or leave it.

Sean Harper 34:35

Same, same. Yeah, you should you should have. Um, and and and it's really a difficult user experience challenge because it's pretty complicated. And there are a lot of variables, right? So it's like you want to expose the user to some of the complexity, but not all of the complexity. So the way the way that the that has traditionally been handled is very ad hoc because it's actually each individual insurance agent who are independent contractors with their own policies and their own procedures, or not, deciding, right? And making a recommendation and really steering you into a product that may or may not be appropriate for you. Right. So we we think we've come up with a better mousetrap with respect to that that involves being presented with the options in a really clear way digitally and being able to fiddle with the dials and see sort of the impacts of well, if I buy this coverage, how does that impact the price? Um and then we couple it with a really good, you sort of digitally enabled customer experience customer support experience too. Because like sometimes you do just have a question, right? So you want to have somebody like, hey, what does this mean? And they'll say, okay, cool, here. I'll explain it to you. And then by the way, I'm gonna make some tweaks to your policy, or you should reload your browser and look at it and see what you think, right? You sort of iterate around the customization of the policy that makes sense for you as the consumer.

Emmalyn Shaw 36:02

Hey, Sean, can I can I just give just for a moment because obviously we're in we're in the the thick of everything, AI, particularly with generative AI. And as we discussed a number of times, you guys have always been very data focused, um, you know, leveraging a lot of the earlier technology with you know, ML, et cetera. I'm I'm curious, but one thing you did early on was really streamlined quoting, coverage advice, leveraging generative AI in that context, as well as other features. What do you, how has that evolved, and what do you think is next in terms of automation at Kin as we think about the entire kind of delivery opportunity?

Reinsurance Shock And Recovery

Sean Harper 36:40

Yeah, we're so machines. Ever since we started Kin, machines have been really good at numbers. And what's different in the last two years is machines have also gotten really good at words. And we're applying that a little bit, customer facing. But really, where that has the most application is in the back office. So you think about the claim, right? We talked about the consumer-facing part of the claim. And that's part of it, right? It's like, oh, there's a claim, we need to like figure out what the damage is, calculate the amount, communicate it to the user. There's a ton of document, because this is a heavily regulated process.

Julie Verhage-Greenberg 37:18

Right?

Sean Harper 37:18

So you actually have a bunch of things where it's like, oh, the claims adjuster needs to write a letter that has this information in it and does not have this information, and needs to have this number and not that number, and needs to be sent to the customer 14 days, but no, no later. Right? There's all these like little nuances around the documentation. Okay, the machine could be really good at that, right? Especially when you have the machine wrapping around a little bit of human judgment. So that's that's what we're using the LLMs for the most right now. It's just automation of some of these like back office things that involve words and language, and it is having a big impact. You know, like um in Q2, we announced our Q2 results a couple months ago. We grew our revenue 30%, and we grew our GNA cost 1%. Um, and that just shows how much leverage you're getting out of this stuff, right? It's not it's not just the numbers anymore, it's now the words. Now the customer facing stuff a little bit, right? There's certain types of phone calls, certain types of customer communication where you can have a machine do that. But especially in this context, you want to be really careful around it. First, because there's a regulatory layer, right? There are certain things where in the law, the law was written about a person doing a thing. Okay, it's not clear does it work if a machine does the thing, right? So that's something you gotta take into account. There's some things where you have to have a person doing the thing. The other is like actually a lot of what we do involves human connection and empathy and trust. And machines aren't that good at that. Like, is the machine good at answering the question? Yes. Is the machine good at you know, providing you reassurance when something terrible has happened? No, not actually all that good at it. Um, so trying to be be smart about what we use the machine for on the customer-facing side, there's some stuff there, and it'll get to be more and more over time, right? As humans become more comfortable interacting with machines, but I don't think there will ever be like on our customer-facing side, there'll always be some human element to it, I think. Because yeah, it's just a high-stakes thing. Yeah.

Julie Verhage-Greenberg 39:28

Something you mentioned there too that we haven't talked about yet is the reinsurance market, too. Like, I feel like those markets have been pretty brutal lately. Um, but Emma's telling me that you guys have been doing pretty well there. Like, what does that sort of signal um for the future as well?

Sean Harper 39:43

Yeah, well, so reinsurance markets got pretty dislocated um sort of the summer of 2022.

Julie Verhage-Greenberg 39:54

Okay. Why was it like what in particular happened around that? Is that just like we started picking up more natural disasters? Or what was the reason?

What’s Next: Cost Compression And Cross-Sell

Sean Harper 40:02

Yeah, a little bit natural disasters, but actually the bigger impacts were inflation and the end of ZERP. And so if you think about it, the raw ingredient for reinsurance is money. And you had a lot of money come out of the system when interest rates went up. Right. So you had all these reinsurers, they have fixed income portfolios. The size of those portfolios actually shrunk, right? Because you have to mark to market the bonds. And so you had a decrease in supply of reinsurance at the same time, almost everything that you're insuring or reinsuring is based to a dollar value. So if you have inflation that's 10% a year for a couple of years in a row, now you have an increase in demand. So you have an inelastic product, big increase in demand, decrease in supply, that's going to lead to a price increase. So even if there hadn't been any natural disasters, you would have seen a pretty big spike in reinsurance costs just because of the supply of money. Now that's that's been digested. So reinsurance markets have have pretty much digested that and are actually now going down a little bit. And that's normal, that's actually a pretty normal cycle. Like you have that um throughout history, you have these sort of like spike up and then melt down. But for a lot of us, those of us who are who are starting something new in that time period, it was very shocking because you went from a world where there was plentiful reinsurance in 2020, 2019, 2020, 2021, and then 2022, you run straight into a brick wall, and now you have very expensive reinsurance. Um that's that's that's all kind of smoothed out a little bit. But for us, it was always like like that was an increase in the in in our input costs, right? Because it's like like if we're making hamburgers, the price of the beef is reinsurance. And if the price of the beef went up, well, we have to increase the price of the hamburgers. Um now interestingly, for insurance underwriters, that was like that cost shock is sort of negative because your costs go up before you can pass the the price on to the customer. There's like a this timing there. For the insurance distributors, it's actually really good because their revenue per customer went up and they're not exposed to the risk, right? So for us, we sort of sit in both parts of this. Like our economics are mostly those of an insurance distributor. We do run these sort of nonprofit insurance exchanges on behalf of our customers. We have sort of like a natural hedge in the business, which was pretty nice and a structural advantage for us. The reinsurers um are also like, I think when the capital is really plentiful and um like like they don't differentiate that much between their counterparties. They're just like, okay, put money to work. It's a beta play, right? We're gonna put money to work, we're gonna put money to work in the market, and they're not as discerning about who their counterparties were.

Julie Verhage-Greenberg 42:53

Right?

Sean Harper 42:53

So when things get a little bit tighter, then that actually is a good environment for us because our underwriting has been really great. And like, for example, with Hurricane Milton a couple of years ago, um, the the reinsurers they use these catastrophe models to sort of predict how much damage will happen to a given portfolio. The catastrophe models predicted a result for Hurricane Milton that was about four times what the actual result was.

Julie Verhage-Greenberg 43:19

Right?

Sean Harper 43:19

So we're creating very differential results because our pricing is a lot more accurate and our underwriting is a lot better. As the reinsurers like that, and over time they've come to appreciate that.

Emmalyn Shaw 43:30

Yeah, I think I think you guys have benefited with that visibility with much better pricing, et cetera, as a result. But can we can we talk, do we take a forward-looking stance really quickly? And then we are gonna ask you just a couple of like more personal questions. But on the on the forward-looking set, and again, recognizing there's always so much, but whatever, whatever you feel comfortable sharing with us, Sean, um, you've hit some major milestones, right? 240,000 plus policies, 100 billion of insured value, you're profitable. What does the next chapter look like? Are we thinking when we think about prioritization, is it um new more states, is it new products, is it embedded? Like how do we how should we think about some of the the next set of priorities for growth?

Julie Verhage-Greenberg 44:10

Which quarter of 2026 is the IPO, is what we're asking about.

Sean Harper 44:15

Hey, hey, hey. Yeah, I mean we we're we in all likelihood will not go public in 2026.

Julie Verhage-Greenberg 44:21

Um, okay. All right, we're gonna have to have you back on then. All right, deal.

Sean Harper 44:25

I'll I'll be here for it. Um, we're really the thing that we're really excited about right now uh is the the cost compression stuff that we talked about earlier. Just get better and better and better at manufacturing this product efficiently. And on top of that, well, our differentiation will always be about understanding the home the best. The folks who live in homes, right? We have this trusted, right? Our competitors, they typically, if you look at the review sites, have two stars, something like that, two out of five. We have like 4.7, 4.8, 4.9, right? So we have a direct, trusted relationship with these customers who, because they're homeowners, are demographically the best customers for almost everything else, right? They have more equity, they have more assets, they have higher credit scores. And so for us, the next thing is like, can we start selling other products to these customers? Acknowledging that we're not going to have the same competitive mode around those products, right? They're not as interesting a differentiation opportunity. But our customers are asking us, hey, could I get from you, could I get a mortgage from you guys? Could I get auto insurance from you guys? Um, could I get a home equity loan for you? And these things tie together in a really interesting way, where sometimes we're actually telling the customer, like, hey, you should do X, Y, and Z to your house. Right? If you did this to your house, it might cost you $5,000, but it'll save you $500 a year on your insurance. And it'll stop a bunch of bad stuff from happening to you that would suck for you. Okay, the payback on that is only okay, right?

Julie Verhage-Greenberg 46:10

Right?

Rapid Fire: Personal Side Of The Founder

Sean Harper 46:10

It's like $500, $5,000 investment, $500 return. Okay, well, what if we could couple that with a home equity loan? Right? What if we could actually um couple that with home equity loan and pro and connect you with a service provider who can do it for you at a discounted rate? Because we're buying in volume. And so here we start to connect the dots between like right now, we're really good at helping humanity adapt to climate change because we're able to be open in places because we know which homes are safer and which homes are less. The next chapter is let's start selling more products to our customers, especially in ways that enable us to help customers actually make their homes safer than they would otherwise be. Make the investment.

Emmalyn Shaw 46:56

That's reinforcing. That's right. Well, what's amazing too is I mean, you know, we talk a lot about so many of these other JSON products. And this notion of like, what is the other kind of underlying data sets that haven't been exposed that honestly with it could add so much more insight, better pricing, risk-adjusted underwriting, et cetera. And I think we sit on something that's pretty, pretty impactful given the history and duration with each customer and some of that kind of corresponding data. So pretty exciting, actually, um for sure, in terms of the types of additional visibility that you provide for some of these partners.

Julie Verhage-Greenberg 47:32

I think it's also a very different route than what other insurance providers, insurance startups have taken, where they've added different types of insurance, whereas you're looking at like you're still staying with mortgage, but like different aspects of the mortgage, like you said, the the HELOC, et cetera.

Sean Harper 47:48

Yeah, we want to be providing things that are customers.

Julie Verhage-Greenberg 47:54

Right.

Sean Harper 47:54

And they don't need to be insurance products.

Julie Verhage-Greenberg 47:56

We mentioned that my daughter will not be making an appearance on this podcast since she's at school, but you you sort of envision by the time she buys a house in like, you know, 2040 something, that it will be a little bit more of like a one-stop shop for all of the home things when it comes to insurance and stuff like that than what it would be today.

Sean Harper 48:13

It may be. It may be. You know, we always have there's always this sort of like unbundling and the rebundling in different directions. You see that happen in a different areas of the economy. Um, so we'll we'll see. Uh, you know, it may be that things for us right now, you know, we have uh an opportunity to make our customers' life easier by bundling more things together.

Julie Verhage-Greenberg 48:38

Very cool. All right, are you ready for rapid fire, and yes? All right, uh you go first.

Emmalyn Shaw 48:44

You go first. Me?

Julie Verhage-Greenberg 48:46

Go first. Okay. You, you, you, you. M go first.

Emmalyn Shaw 48:49

All right, let's see. Okay. Uh, if you could have dinner with anyone, past, present, maybe not future. Let's do past and present. Who would it be? And maybe why?

Sean Harper 48:59

Uh Milton Friedman.

Julie Verhage-Greenberg 49:03

Okay. Okay.

Sean Harper 49:05

He had a big impact on me. I was really interested in economics when I was in high school, and I read a lot of Milton Friedman, and I think he has uh pretty interesting um impact on on how things have gone, right, with the markets. Right. Markets are like I think I I think like the trading with each other is like a fundamentally human thing. And um when we deviate from that and start to create a lot of restrictions around trading with each other, it has a lot of unintended side effects. I think I think Milton Friedman was one of the you know most important voices to call that out. Super interesting guy.

Julie Verhage-Greenberg 49:49

There you go. All right. Um, personal highlight from 2025. So not related to Ken. Anything personal that happened this year that um was a highlight for you.

Sean Harper 50:00

I had a big one. I got married this summer, so that was exciting.

Julie Verhage-Greenberg 50:04

That is a big one. Congratulations.

Emmalyn Shaw 50:07

Yes, very exciting. All right, last one. I think I know the answer, I'm gonna ask anyway. How do you de-stress?

Sean Harper 50:15

Um, I don't know. I hang out with my talks.

Julie Verhage-Greenberg 50:19

I was gonna say it's gotta be related to the fur babies for sure.

Sean Harper 50:25

Um yeah, I don't uh that's I I I don't know. I just do normal stuff. I just like work out. Those were the two I was gonna guess.

Julie Verhage-Greenberg 50:35

What's your workout of choice?

Sean Harper 50:38

Um, I like I like uh I do CrossFit. I like I like anything that involves um moving intermediately heavy weights very quickly.

Julie Verhage-Greenberg 50:50

Okay. All right. CrossFit is a good one for you then. Have you ever done any of like the high rocks uh tournaments or anything or no?

Sean Harper 50:58

I've never done high rocks. My wife has done it a couple times.

Julie Verhage-Greenberg 51:01

Oh, interesting. Did you guys meet through working out or anything like that? How'd you guys meet a working out? There you go. I love it. Very cool. I love that. See, we my husband and I met on a dating app, and then the first time I taught fitness classes for a long time at Equinox, and I made him come to one of my spin classes, and he was extremely hungover the first day, too. He's like, oh my God, I still have to like crush this front row and everything. Um, and it was one of those cycling classes where like there's a leaderboard and everything too. So like I knew how he was doing. I was like, why did you get like he actually he's still a woman? He felt horrendous afterwards, and he's like, I thought I was gonna throw up, but like he pushed through it. I was like, You dummy, you knew I felt like I sprung this on you. You knew you had to go.

Emmalyn Shaw 51:45

I was impressed that he showed up. That's right. You were worth it though. Good for him.

Julie Verhage-Greenberg 51:50

Exactly. Exactly. And now there's Quinn, so it's okay.

Emmalyn Shaw 51:56

Amazing.

Julie Verhage-Greenberg 51:56

Well, thank you so much, Sean. This was great.

Emmalyn Shaw 51:59

Thank you so much. So great to spend time with you.