Lemonade Car Deep Dive: Michal Langer Interview

Hi, good to be here.

So it's a combination of things as you mentioned but I think there are inherent benefits to the fact that we're out of the bat a telematics product. So everything is weaved with telematics in how we approach things if it's pricing and underwriting and automation and our experience. So I think the high adoption is an artifact of how we built telematics. So I'll give you some examples. We don't have a customer base that we now need to they bought one thing and now we need to convince them of the benefits of telematics and moving to giving us data. We're super comfortable with the product we've built. We believe it's better for our customers to have telematics. So we're up front about it and customers who buy Lemonade car, they know that they're going to share their driving data with us. We also I think other companies have multiple apps. So they have their main app and then there's another app. They need to convince the customers to download from the third party provider. And we have with us we have our one Lemonade app which is I think beautiful and easy to use and our customers use it. So they just, you know, turn on location services and that's about it.

I think on top of that, we're also we are weaving it into all of the experience. We talk about why it's beneficial for you and I think it's helpful that it's true. You do get better prices. Your claims are going to be managed faster. We have perks. So, for example, we have crash detection. So, if something happens, we reach out. We see that you're okay. We send you emergency services. We have tree planting. So, we help you with the carbon footprint and you can see your family forest in the app. Cute stuff like that. So, I think everything is tailored around this. And on top of that, as you mentioned, you have young customers who are tech-savvy and they feel comfortable with sharing information. So, all of that is driving these gaps.

So, you know, we build a really powerful brand for young consumers at this stage and they're coming to us and our way to get them to convert is giving them better prices than the rest of the companies and that goes through knowing that they drive well. So, it's a key priority for us to be able to price them as soon as they come to us. So, the faster we do that, the better we're going to grow. So, it's a big focus of ours and I can't share too many details because I do feel like this is a secret sauce and we have a unique approach going into this but I can share that we're looking at different things like customer self-assessment or third-party data or predictive models based on all of the information that we do know at the point of sale. And then after that we're also looking into how fast once you start driving with us we're able to know how you're driving and with what confidence and when we update your price. So all of these things are right now in tons of testing and we're seeing really great progress and I'll share probably later on when we're feeling more comfortable with sharing secret sauce.

Yeah. So these things there are advancements all the time but I can tell you if you're thinking about driver versus passenger or it's a statistical model that you know spits out the probability that this is the driver or the passenger it's not 100% correct. So you could have instances where you're mixed up, but it's not enough to really make a difference in your pricing. We're looking for consistent patterns on your driving behavior that are going to tell us what type of driver you are. So if you know somebody has their daughter's boyfriend driving the car for one time, it's not going to make the difference. So there are a lot of different things that go into it.

Yeah, I know there's a lot of interest in this area. So maybe I'll try to answer something a bit broad. So there are a lot of advancements in technology for collecting telematics data. There are different companies and providers and they get advanced all the time. And that's not the focus for us. For us we just care about getting the data. The magic at Lemonade happens with what when the data arrives. So smoothing it out, using it for all the predictive models, understanding how you're driving, weaving it into our automation, those type of things. That's kind of where we're focused. So what we invested in is creating a system that's super flexible and agnostic that whatever comes in, we're able to ingest it. So as these things advance, whatever is best, we're going to absorb it and do whatever we want with it. And what's driving our decisions are assuming you have enough signal that you need for all these things. What's the best experience for our customers so it's seamless and it doesn't bug them and cost. Ideally we don't pay for it and the customer doesn't feel it and we get all the data. We don't have a solution that solves all of those yet. So we're experimenting with different options and each has advantages and disadvantages. So we have in some places a device as you mentioned in some places a phone. We have places where people have the device and the phone together. So we can see we can learn about the differences between them. So specifically a device it's better at accurate mileage collection and you just plug it in and you forget about it but then it has a cost and you need to ship it and the phone is better at understanding handling a phone or a distracted driving and you just turn on location services but then there are system notifications. So there are pros and cons to both of them and we keep tracking and improving and using whatever is available but at the end we want to either have just a drop down when you purchase and you can use whatever feels more comfortable for you. Or there's one thing that will triumph all the other ways. You know there are cameras, there's connecting directly to the car. So we're keeping track of everything until we have the perfect solution.

Yes, some states are stricter than others. And you know, for example, if we charge you by the mile, we need to prove that it's really accurate. So you know you might there we prove that it's accurate and we share a lot of data and depends on the states but yeah.

So we're using different providers to get the data the telematic data and then once it got to us we're building the models to predict driving and everything because you want to connect everything to your claims and your customers and that's how you get to most accuracy. So that's the direction we're going with.

That is correct. So in California, we're able to use telematics in order to change your price based on how much you drive, which is still a really big indicator. It's basically predicting exposure and risk. But not based on how you drive. Maybe one day it will change. I hope we can prove why this is really beneficial for our customers. But that's just a part of the job we need to navigate the restrictions in each state. In general I think what you're alluding to is the fact that we have in different states we have different programs that we're learning from. We have two constructs or payment types. One is pay-per-mile or pay-as-you-go. So we every month we'll charge you based on how much you've driven and then we have kind of a fixed price which is you just purchase the policy for 6 months. You can pay every month or every six months and it will be fixed.

So maybe those are the things you're seeing that might be different between different states sometimes.

So here we have two guiding principles when we get to when are we going to change or identify changes in how you're driving or the pricing for it. So one is we want to make sure that we're as flexible as possible and we can change anything at any time. And then the second is the experience. So there's a lot to navigate there from a flexibility perspective. We're able to recalculate at any point, make any changes that we want. And we also invested a lot to create more structure or degrees of freedom in our filings which has been a challenge in the beginning and now we can test and do a lot of things. Where we are experimenting now and trying to find that optimal point is once we know how you drive when do we make a change and we're experimenting there so that largely it's at renewal but we're really trying different ways and points in time where we can update sooner and I think there it's really how do we create enough peace of mind for our customers but also letting them realize the benefits of them driving well. Because even we've seen that sometimes you can reduce the price for someone but not everyone's comfortable with change or predictability. So you change once and then they think well is it going to happen again and some people just prefer to have you know okay this is the price that's it. So, we're really investing in setting expectations and messaging why and when and being upfront and kind of finding that right place where we're getting the best prices for our customers, best loss ratio for us and also that you know they're not stressed about it and it's not something that changes for them all the time.

So there's the per-mile and then there's the fixed one. So with the fixed we'll update at renewal based on how you drive, but also how much you drive also makes a lot of impact because somebody that drives a lot or doesn't drive at all it does change the telematics score as well. So at renewal, we will update both of them. So, in any case, if you're a person who doesn't drive a lot, you're probably going to get a lower price, right? And with the fixed, as I mentioned, we're trying to move it earlier. Just because we already know you should get a better price, so we want to update it for you.

Okay. So obviously this is an area that's very interesting for us and we're following it very closely. I think the change here is not going to be a binary change, right? It's going to be something gradual. There's going to be a certain percentage that the car is driving versus the person is driving. And so I think innovation here is going to be in the form of lower prices for the percentage that you're not driving or different structures or coverages. So I think in that sense we're really set up well for innovating in this area because we're really focused on pricing really well every single mile whether that mile is driven by a car or by a human. So we'll just know that by a car it's driving safer so it will get a lower price and you know we're very open to changes and we can change whatever we want at any point. So, we're working on future integrations that are going to help us understand every autonomous mile better and how it's progressing. So definitely something we're tracking and excited about. I will just say that, you know, it's hard to predict anything about the future. Who knows? Right now it looks like it's going to take a while but even let's say things get so much better and it's more prevalent and TAM decreases by 10% or 20% in next few years maybe it's something that Progressive and Geico will be concerned about but for us like car is still going to be infinite we can 10x and then we can 10x again so it's still it's not going to touch us anytime soon so by then I think we'll be ready.

Yeah. You're just going to have a super safe driver and all of the rest of the things that we're doing as you mentioned we're going to streamline down.

[Laughter] I can give hints. Okay. It's a combination of a lot of things. So there's state availability and marketing efficiency and experience and pricing adjustments. So let me give a bit of detail on each. So we're continuing to roll out more states for existing customers. So they just add them as soon as we get there. We launched Colorado. We just launched Indiana. I don't know if you've seen it. And with our experience we've really been focusing on how we speak to our customers, what resonates, how do we talk about car, we mapped every single place you can possibly reach out to customers, removed friction. So in general, the more customers we have, the better we get at predicting and reaching out to them. And our all of our marketing efficiency and machines are kind of finding their place and things that work for us in renters and pet we get into the maturity that can go into car. So all of that is really kind of shaping up and it's just going to get better the more customers we have.

And then there's well first existing customers that are adding car their unit economics are insane. So we don't have cost of acquisition which in car is amazing. They have lower loss ratio higher retention. We see almost no new business penalty in car which is crazy. So we have a lot to play with. So, we're giving them more discounts, which is obviously very helpful to get them to buy. And then in addition to that, we touched on this a little bit before, but we know a lot of things about our existing customers, so we can predict who from these renters or pet customers can be a really great car customer. And then we're able to give them even more discounts. So, we have customers that are existing and we know that they're going to be good that they're getting amazing prices. So, that's very very helpful and you know if anyone's watching this and you have Lemonade and the car is available in your state and you're not checking us out, it's really like I just feel bad because it's money on the table so you better, you know, just check.

Yeah. Where you're covering every single possible reason to back. I'll tell you this car is a wild horse. We can you know let it go and let it explode but we want to do it really carefully because it's a massive product and it can also do a lot of damage. So we're being very thoughtful about it. It's a strategic product for us. We're building a multi-billion dollar business and for us not that it's definitely not that we don't have a sense of urgency. I've never been accused of being patient. But a quarter here or a quarter there, that's not what's going to be kind of the that's not the focus for us. So there are a lot of things that we need to prioritize. There's so much to do. Each state we're looking at, okay, what's the regulatory environment? Can we use the advantages that we have? How fast can we change? How many customers we have? And then also rolling out in different states. We're also mapping that versus working on existing things. So you know we I've shared a bunch of experiments we're doing and we're seeing a lot of things that are working and then you know you see something that's working and you're swarming in and then you keep iterating and swarming in. So, if it's taking us some time to kind of catch up with inflation, get to the right rates, then get into the most sophisticated pricing and segmentation, and now everything is really working well together, and we're seeing tons of really great stuff, but we want to do it right. So I think we've shared in the past that 2025 is kind of the feature complete year and 2026 we're going to realize more of that acceleration. And so far things are looking good. So, car will be big, but we want to make sure that we're keeping everyone safe.

Okay. I have a lot to say about this. That's something that's very near and dear to my heart. Okay. So, pet is an incredible product. It's the perfect product. We've talked about this in the past. It has high frequency, low severity, so you can learn really fast. And when we have data, we know what to do with it. And we're doing really amazing things. And for me, that's kind of like a pet is like a glimpse into the future or you know, kind of proving our thesis and you can see what happens when we have enough data. And to me, it gives a lot of confidence about what we are able to do with car. And you know I should know because I built our pet insurance from scratch before I moved to car. So car is much more complicated and there are a lot of different things that go into automating car. But also the reward is very big. So everything that you do, you're removing massive things and already we're seeing really amazing things in just the way we rolled out car and the things we've built so far even before getting to all the things that we have in the future. So the way we approached car at the beginning we really sat with industry experts and people from all of the big companies and we mapped every step of the process and here you have hundreds of people and then you lose all these days and then you need to call them again because you miss something and then somebody goes in and says if this then that and they make a decision and just the way we built it is you've removed all those hundreds of people from scratch. You're making all these decisions automatically. So, that's kind of how we rolled out and already we're seeing our LE or loss adjustment expenses are up to par with the biggest and most sophisticated companies in the industry and we have a bunch of metrics that by the end of 2025 are going to be best-in-class and then a bunch of them by the end of 26 best-in-class. So, I think it's pretty amazing that we're already in this place. And it's you know, we have nothing. We don't have buying power. We don't have economy of scale. We don't have all of the automation that's coming. So maybe I'll give you some examples for the things that we have going for us now. So, 80% of our car claims are through the app. So that's an amazing number. It's kind of the opposite than what's happening in the industry where the majority of the claims are through the phone. So we have all of that data coming in validated. There are no humans. We can start automating a lot of the decisions and streamlining the processes. We have for example people something happens they open the app and we can show them the shops that they have near them that are internal shops. And it's important because we vet these shops so they have good experience but we're integrated with them. So, we save a lot of money on people going to these shops and it's something that other companies are really working hard to get that into high percentage and we basically just have that percentage super high from the beginning. Then on top of that, you layer telematics. So we know what happened like we know when the crash happened, the speed, the severity or where the hit was, where were you? So, we can identify fraud. There are a lot of things that we just know based on telematics like we can understand if there were total loss or who's at fault sometimes so that allows us to streamline a lot of things so this is you know I think if we you and I spoke a year ago I would say and we have a road map with tons of things and automation but with the things that happened in the past few months even that's been you know blowing our mind and we're very soon we're going to have AI agents that are verifying things and following up and summarizing and assessing. So I think there's no limit to what we can do with car. It's very very exciting for us and we're touching every cost layer with you know less humans, less time, better experience, less fraud. So feeling very very good about car automation.

You have contracts and you don't have fraud. But you know mainly the process is streamlined and it's automated so we don't have humans involved. So that obviously saves us costs. Things are faster and in general the reason we recommend our customers is because we know these are good shops that are going to give them the best experience but also we're going to manage the claims so much faster so it's better for them. So it's good for everyone which is why we're recommending it but obviously saves us costs.

So Lemonade I think when we have data we know what to do with it and we're really really good at it. I think that's something I feel comfortable saying and I keep repeating it and but it really drives everything for us. And getting data is painful. It takes time which we always feel like we don't have and you need to lose money in order to get it because you know claims and loss ratio. So we're always in this search of finding data faster and using data faster or learning faster. And I think a lot of the messaging that we had around the acquisition of Metromile was a collapsing time and that was the reason for us to look for an acquisition like that and Metromile was a wonderful opportunity for us. It's a telematics company. They had almost half a billion trips and billions of miles tons of claims. So we took all of that data and it's used for everything for our predictive modeling for underwriting experience. We were really able to use it to boost everything that we have. And on top of that we had you know we got customers and IFB and the 49 licenses. So that was just like a cherry on the cake and the terms that we had it was almost like free. So I think overall very very good in terms of the unit economics and the learnings and it gave us a big boost. We did have you know if I'm kind of sharing the whole learning or the experience from the acquisition we had a moment there where we kind of needed to make a decision between moving fast and having multiple systems that don't necessarily talk to each other or keeping legacy and those type of things versus taking the time and the resources to move to one streamlined system and everything is advanced and agile like we believe we want to have and it's I think it was there was something symbolic about that decision because it's kind of like what company do we want to be and I think it's so I have so much empathy to companies who choose to just you know just move forward and not take the time to do it because it's painful and it takes time and you need to kind of get patience from everyone around. But I'm so happy that we've done that and that we're now like I'm very proud of that decision. Now we're running everything is speaking to each other. We have you know like we always say we have one system and everything is connected. So, now it's super fast and we're in good shape.

Oh, they've been moved over. Yeah. Everything through the system of Lemonade, the data of Lemonade.

Yeah. I think it's true to say we obsess about every single time we speak to our customers about anything. So, that was carefully thought as well.

We had a lot of bumps and there are a lot of puns in the car. So I think a few things that we had to learn in car. One thing is when I don't know if you remember this but basically when we launched car inflation went insane like you look 10 years before that there was nothing of that magnitude you couldn't even so we basically just launched car and this crazy inflation happened and we had to kind of chase rate in a really tough regulatory environment which is car. So we had to you know up our game in filings and managing rate really really fast. So I think that was one thing just entering the car market was an interesting experience. I think in general just the wheels of time in car like by the time that we see something want to change understand something and the time that it takes to actually make it from a regulatory environment it's been a learning for us and I think we've invested a lot once we picked up on it on how we create structure within filings and programs and tools to allow us to change things and now I think we're a much much better place and we're able to change and test and we've created kind of that structure for ourselves. We talked about Metromile. I think there was a lot of learnings and kind of defining moments for us with that acquisition that I think we're in a really good place at the other side of it. And I think in general in car, it's true for all of the products at Lemonade, but in car specifically, it's especially true that every word matters. Like we really really see a huge difference. It's such a competitive and complicated product that you can see, you know, in an email we try different phrasing and it can be completely different outputs or you can have people tell you listen it's been tried in other places 20 times it just won't work and then we find the angle and then it does work. So really really understanding how we use experience to drive action. It works and it's worth the investment and it's something that I think we really invested in in car and learned. I can tell you one or two launch stories if you want. I don't know if it's interesting. One thing is you know you asked me about the telematics adoption and when we launched car telematics was kind of you know we thought it's going to be so amazing and everything is centered around it but we weren't sure are people actually going to give us their driving data and we were kind of we know if it doesn't work we're then we need to go back to the drawing board like what if we roll out and there's nothing and you know we planned like tons of like very long roadmap of improvements and we had internal bets and there were very low bets of what would be the percentage and then we rolled it out and it was just super high. We were like, you know, that doesn't happen. Usually something happens and then you need to iterate and it was just people just got it and we still had to work on it and improve because you know we're never satisfied. But it was you know I remember we kind of took this breath and said okay I think this is going to work and so in general maybe I'll end with another story about car or how we planned car after we finished pet and we felt we're ready we came into car with a lot of healthy fear as you should and we kind of thought and planned on how we're going to get into the market what's going to be the advantage how do we approach this and I had these experts across Lemonade who've been in the industry for many many years in claims and insurance and legal and finance. And we would come to them together or separately. And kind of show, okay, here's how we're planning to do this and I would say, listen, tear this apart. Like, tell me all the reasons it's going to fail. And they would and they would say, ah, this is you're going to fail. It's not going to work. And then I would go back to the drawing board and then come back again. Okay, how about this? and like back and forth again and again until I remember one time I came and I showed the latest to one of the experts that we have in claims who I think very highly of and he looked at it and I remember he kind of paused and he said, 'Oh my god, we're going to win this.' And I remember I didn't even need to hear him say it. I saw it in his eyes. I saw like this thing. Okay, this is going to work. So for me that's kind of the story of car in general. I think we're not afraid of what can go wrong or what's a disadvantage. We're excited about understanding it and turning it into an advantage. And car is inevitable like we have such a good product which our customers already really like and we have telematics and technology and automation and existing customers. It's just it's all going in the right direction and I think we're just we're going to be relentless and thoughtful about it until it's just going to take over. So yeah, it's very very interesting and exciting to work on car and I think the rest of the team feels that way too.

Thank you. Anytime.