The Future of InsureTech: What InsurTech will look like in 2030 with Metromile CEO Dan Preston

Count us down from 3, 2, and 1. Hello everyone, welcome to SaaS for Money. We've had such an incredible morning. Thank you for joining us at this incredible digital mega event we've been having. This session is all, is all about the future of InsurTech, what InsurTech will look like in 2030. Just a few housekeeping notes before I pass it over to Dan. He's the CEO of Metromile. He'll be your speaker today. Drop into the chat where you're joining us from. I know SAS has got a really big global audience. We love to see from around the globe where you're tuning in from. You can submit questions to Dan. We'll have plenty of time for Q&A since this is a 50-minute session. So feel free to drop into the Q&A or into the chat or your livestream, wherever you're tuning in. We'll aggregate those. And we'll have time for questions at the end. As a reminder, as always, these sessions are always recorded and available post-event. We're actually live streaming, so you can see this on our FASTR YouTube channels, and we'll share the slides with everyone afterwards. So without further ado, I'm going to pass it over to Dan Preston. He's the CEO of Metromile, and we're so excited to have him here today.

Fantastic. Thanks, Melia. I'm really excited to be here. Um, and so I'll, I'll start by maybe just introducing myself real quick. So my name is Dan. I'm the CEO of Metromile. My background is largely as a data scientist. I started my career actually in astrophysics, applying machine learning there, and ultimately started a company in payments, sold the company Intuit in 2012, and then actually joined Metromile early on as the CTO when we were a very small company. And became CEO about a year after that. It's been really exciting to see how data science, what largely my early part of my career has been built on, evolved to become what is now transforming insurance across really every line of insurance overall. So I thought it would be fun to start by sharing a little bit of the foundations of insurance going back hundreds of years. Talk a little bit about how it started, why, because I think it teaches us a lot about where it's headed as technology advances forward. So insurance largely began in the 14th century. So in Italy, many of the ships that were transferring cargo to different parts of the world were starting to go missing. And so the early insurance contracts were created in order to mitigate that risk and spread the risk of losing cargo or ships across all of the fleet. What started to happen over time, however, is they started to find that certain ships or certain captains or even certain routes would have a higher chance of losing cargo. And because of that, we started to then adjust the prices based on those different— based on those different characteristics. I move forward one. Um, and so over time, insurance in many ways has become a critical foundation for resilience between different communities. They think of, for instance, different groups of farmers or different professions as those communities look to, to spread risk out across to mitigate the risk of individual events, you then would set a price for that group of customers or group of individuals based on their propensity to get into some type of situation that would then have to be covered by that insurance risk. Um, if you want to move forward one. And so like many industries, as things have moved forward, the advancements that we've seen have moved away from these kind of community-based risk pools and towards ones that are more demographic-based. And so much of the advancements we've seen in the last, you know, 20, 30 years have largely been based on the use of things like credit or selling products directly to customers, as opposed to within those communities that support one another through these types of products. And this worked well. Over the last many decades, credit has actually turned out to be a really important way that some insurers, a small set of them, actually have been able to take advantage of a piece of information that others were not pricing with, effectively taking away the low-risk customers and then forcing the other insurance companies to have to raise their rates for the others who are then left over, over time. And the problem with this, the challenge over time, is that what we're left with is a, is a very transactional experience with insurers because it's based on these proxies, say, less of who you the individual are and more of, say, your credit score, your occupation, etc. This has become more of a transactional experience with these different insurers as opposed to one that is about a community of people who are looking to ultimately support one another, um, to mitigate that kind of risk. And at the same time, because those advancements have largely, um, still relied on the same type of operations that have existed for many decades, there's an enormous amount of operational inefficiency embedded within insurance organizations today. So think of about half of what you pay in your insurance premium is actually there to cover the cost of that risk. The rest of that is going to things like fraud or the cost of managing claims in the first place, or in a lot of cases, a very high cost load for the distribution of those products. Just the advertising that goes into awareness for commodity— commoditized products leads to higher premiums for customers over time as well. So what you're left with is fundamentally a product that is not personalized to your benefit, and is also leading to a, a high cost to the customer at the end as well. And so this, this proxy-based view of risk is leading to this world where you have a large group of customers who are overpaying for, for their insurance for a lot of reasons, whether that's being priced based on these proxies or groupings of customers, or because you have this high operational load that is, um, there's— there are the inefficiencies of insurance companies today that aren't taking advantage of newer technology. So as the world moves forward and we start to see the technology benefits in, in all industries, we're seeing this happen in insurance. And if you look across, uh, insurance overall, fundamentally is about data. And so we're seeing the access to data increasing dramatically over time, in particular the way that we sense the world. And so as the world digitizes very rapidly, we see sensors on every person, every car, every moving thing across the world increasing exponentially. And as we start to capture more and more of this data, we're now able to understand the physical world in a way that we weren't able to before. Um, and so as the physical world is now measured in detail, the uncertainty of what's going on declines of our time. And so if you think of the fundamental job of the insurance company to understand what is the risk of an event happening, instead of looking across as a whole pool, we can understand what is the risk of that individual event or the individual person based on how the car or whatever it is that might be being used. As an example of this, that moment in time that is so critical for, say, an auto insurer is when you get into a car accident. And so if we have the high resolution to be able to say, we know what happens at the time of the accident because the sensor data that we're capturing from the vehicle can tell us what part of the car was struck, how severe the accident was, etc., the manual work that used to go into having to talk to customers on the phone and ask them what happened, look for police reports, etc., goes away. And you can actually interact with a customer based on what truly happened in real time and without having to have all of this operational load, leading to a better customer experience, but also leading to higher operational efficiency over time as well. And so what's evolving here is that now, as we understand what truly happens, we can price by the moment, by the actual moment of risk, whether it's a mile driven or the day that you use— use an Airbnb or whatever the act is that you're moving at that time, we can price by that and actually offer you a better experience through the use of that sensor data over time. And ultimately, these, these, these understandings of real-time risk over time and what really happens at the accident scene is allowing us then to create fully automated claims fully automated experiences for customers that then lead to, of course, the, the operational efficiency that can then drive improved pricing for customers in the end. What's really exciting about the world that, um, that evolves from this is one where the insurance company and the insured are then truly aligned, um, because we are able to help customers decide better how they can ultimately lower their losses, get in fewer accidents, or experience fewer risky events. Of course, the, the costs for the insurance company are coming down, but also we're allowing the customer to benefit in the improvement in that risk as well by passing on the pricing benefits in real time. So coming back to that foundation of insurance largely being about communities supporting situations where you have risk to spread out across that group. This movement away from the community has eroded trust in insurers and has moved insurers to using more of these proxy-based factors. And these proxy-based factors oftentimes correlate with things that today We are actively trying to move away from to create more mobility and more equitability in financial services overall. And this is not just about companies building a brand around equitability. This is a fundamental restructuring that I think we're seeing across the industry overall. And you see this with some of the regulators now that are asking to remove things like credit score from the way that we price. And ultimately what I think the world is moving towards is one that is based on what the customer's true need is, one that is based on the real-time events that are happening that you can measure with sensor data that allows customers to actually be in control of their outcome. And so the, the sources of data can be from anywhere. It can be from the car itself, it can be from a home, it can be from a wearable that you have on your wrist. Any of these are sensing the physical world and allowing you to actually understand what is truly happening as opposed to some proxy for the group of customers that you may look like. And with that, it allows us to then usher in new types of products that can support different kinds of ownership. So a great example of this in auto insurance is the standard 6-month model for auto insurance. You pay $1,000, say, every year. Doesn't fit when you're constantly switching between renting that car out or using it for ridesharing or whatever you may want to use it for. In that case, we should actually only be charging you for the periods of time when you're actually using that car or when you are driving that on your own personal insurance. Another great example of this is as cars become more and more autonomous, The periods of time when you are driving the car, of course, you may want insurance to cover your own liability. But if the car is driving itself and you're allowed to fall asleep and let the car drive down the highway until you get off, that risk likely is transferred to some other entity. And we should be able to price each mile driven or every moment in time differently based on how the car is being used. And I think the ability for us to measure in real time is allowing us actually to lower the overall cost of ownership of vehicles and other things by being able to differentiate when someone needs insurance and when someone doesn't, and being able to give them more control over what they're paying for it by incentivizing and actually giving rewards back for either driving less or being less risky or safer overall. So I'm looking forward to some of the questions. I hope this gives a really brief overview of insurance over the course of the next 10 years or so. I think there's a lot of really exciting questions that we can walk through. So let's pause there.

OK, awesome. Thanks, Dan. We have some good questions pouring in, and there's more. So thanks, everyone. Keep them coming. The first one is, why do you think InsurTech has exploded with VCs and IPOs recently? It's just like a sudden boom, or what's your take on it?

Yeah, I— so we started Metromile about 10 years ago, I think, actually this month. And at that point, I think there was maybe just us and one other. And then over the course of the next 6 or 7 years, we started to see a lot of new companies start to to start startup. And there's been a fair amount of interest, I think over the last 5 years. It's really been in the last year or two where I think a lot of these companies have developed enough scale to attract really late-stage capital and growth from there. But the answer ultimately is there's an enormous opportunity. There's trillions of dollars of insurance across the world in total market size. And there's a lot of opportunity to apply technology in different ways across many different kinds of— lines of business.

Yeah, that makes sense. It does seem like sort of in line with that, right, that there are like hundreds if not thousands of new insurance startups being minted almost every day. How do you see the market supporting all these different apps that are coming up?

Sorry, you just cut out. Can you repeat that one more time?

Oh yeah, no worries. I said, you know, it seems like there's a lot of newly minted insurance startups being funded and founded almost daily. What's your take on how the market can support this like huge influx of growth?

Yeah, I think in general, I mean, there are dozens of lines of insurance where fundamentally the opportunity here is a bunch of new data that's allowing us to price differently or offer the product in different ways. And we're just at an inflection point now where there's an opportunity to probably go after almost every line of insurance. In that way. But this is an extremely large market. So I expect that insurtech has probably only scratched the surface of the potential scale from here on out. But I think a lot of the investments going behind the idea that there is an enormous amount of scale out there.

Yeah, and for those who are maybe a bit unfamiliar with it, right, because it is sort of a disruptor in the space, how would you describe the inner workings of how a company like MetroMile really works? Obviously, it's sort of like a, you know, a simple pay-per-mile kind of tagline you guys have that you're marketing. But how is it that you can like better segment, you know, your customers based on that marketing versus, you know, you maybe don't want the traditional folks who are buying, you know, calling up their agent to buy insurance?

Absolutely. No. Yeah. So, okay, for Metro Mile, think of it this way. There is there's a linear relationship between number of miles driven and expected losses, right? So for every mile you drive, there is some expected, you know, number of cents of loss that you have at that point. And it's rather intuitive. The more you drive, the more intersections you go through, the more cars you pass, the more likely it is you get into a car accident. But the way that insurance is largely priced today is you get one fixed cost based on a set of demographic variables. So the the person who is a professional that gets charged $1,000 a year, drives 10,000 miles a year, is getting charged the same as somebody who has the exact same characteristics, which are 1,000 miles a year. So the idea here is that if you can charge on a per-mile basis, you can offer substantially lower prices to the people who are not driving as much. But the only way that that's possible is if you can get the true number of miles driven. Because even though your insurance company asks you how much you drive, everyone always gets the lowest tier. But you never get it right.

Everyone always gets it wrong.

Yeah. Yep, so that's the fundamental opportunity. Now, of course, as time goes on, right, like you can use more of that data to better price over time. If you're a safe driver, we can lower your per mile rate, et cetera. So that's fundamentally the starting point. And then from there, you ultimately have the other advantages that accrue over time.

Great. One thing I think, you know, you're sort of springboarding into is like, you guys are a bit of a disruptor, right? And it does take trust to disrupt, like, you know, the folks who are more hard-coded and are used to picking up the phone, calling their agents, and buying insurance. Obviously insurance involves a lot of trust, making sure that wherever you get it, right? Like you guys will be there when something happens, right? Like I'm gonna get a payout, I'm gonna be covered. Like I don't have to sweat this 'cause I have insurance. Like how do you, how would you, what would be your suggestion or advice on how folks judge a new entrant in particular, right? So something, something new like MetroMile, how do you build that trust with people to say like, yes, we're going to be there for you, just because maybe you haven't heard of us yet doesn't mean like, you know, we're not going to be there for you in the times when it matters.

Yeah, we think a lot about this question of trust. I would break this down into 3 ways that we see customers think about this. The first is like the question of solvency, right? Like, Will this business be around for a long time? Insurance companies have a very storied history over the last 100 years. And not all of them have always been there. And so part of this is making sure that the company is well-capitalized, that you have strong reinsurance partners. And so, you know, part of that for Metromile, of course, has always had strong reinsurance partners and well-capitalized through the public markets now. The second is through claims. Claims is basically when you get into an accident, are you going to be there? Both monetarily but also in spirit, like, will you treat me fairly?

Right.

And I think that's something that has been really core to our story of how we aim to not have that kind of adversarial conversation with customers, where instead of us kind of interrogating you about, you know, were you really at the accident scene, et cetera.

Yeah.

We actually can just validate that what you told us really happened, and then you can get that handled in your app within seconds as opposed to having to go through all that. Um, the third part of this, which I think is emerging for a lot of insurance companies, especially the newer entrants, is how are you going to use my data? How is that done fairly? And so one of the things that, that we thought about early on was how do we deliberately build an experience where people see the benefit of data sharing in their day-to-day lives? So we built an experience that tries to go beyond just claims, but helping people avoid parking tickets, diagnose car issues. All that's kind of enabled through the data. Yeah, and our customers love us, and the reason ultimately is we're trying to help people save money in other ways. But I think what this does is it helps to build a bridge for people to understand that our goal in capturing this data in the first place is to build value for the customer, not in selling that or doing anything that would be adversarial to them.

I— okay, I— so I feel like Now, I hear a lot of things that you— I think I'm hearing that you guys use to differentiate yourself not only from all the new people in InsurTech, but obviously the sort of old guard, like obviously ease of use. But it sounds like you guys are— you really think about being proactive. And I know you started as the CTO. Like, how do you program that into there? Like, give us your data? Because I feel like app adoption is so hard to get people to actually, you know, take the 2 seconds to put that in so that you guys can make it a better experience for them.

Yeah, so part of this is just built into the, um, like the basic starting idea for the company where because we charge on a per-mile basis, we need to capture the true number of miles driven. And the way to do that is we send a customer a device, they plug into their car, or we directly connect with a vehicle. Um, and with that, it allows us to charge on a per-mile basis, but it's also the same way that we can for instance, measure the state of your car to help you diagnose car issues, or the location of your car to help you avoid parking tickets. All of that is coming from that same device.

Awesome. Okay, got it. And then I have— we haven't— we have a few more questions. I'm gonna keep, uh, keep them coming at you. How do you see incumbents reacting to the disruption by all these startups, right? We've been chatting about it a bit. Do you think they can still catch up, or are they simply gonna get left behind?

It's a really interesting question. You know, there is an enormous benefit of having scale and the brand that consumers trust day one.

Right.

The challenge, however, is that, and this kind of goes back to the original challenge of like credit score, which allowed some of the names that everyone here would recognize off the top of their head, is that when you have an existing portfolio of business, it's really challenging to introduce some new way of pricing because what happens is you, you lower prices for all of your lowest risk customers. So say your, your average revenue or premium comes down for half of your portfolio. And then once you do that, of course, even though it's probably the right thing to do for those customers, you then have to increase your prices for the other half of your customers, which means they're all going to leave for somebody else. And so it ends up being this kind of cannibalistic issue for insurers where they start to introduce too rapidly some of these newer concepts, it cannibalizes the existing business that allows them to have that scale. And so one of the things I think we're seeing, and I think we've seen historically, is that as some of these newer ideas get introduced, you really need newer entrants to be able to spark the idea and build initial scale.

Yeah, yeah, no, that makes a lot of sense. There's just a few more questions. Here. I know one thing you touched on is that, you know, you were talking about future trends and what InsurTech could look like in the future. Where, like, where do you envision even like as a user, right, of your own product, and I'm sure other, other apps, right? Like, where do you see this going? Like, you talked a bit about like, you know, the Teslas and the self-driving cars, like, where do you envision this going down the road?

Yeah, well, I think for auto insurance in particular, it's going to be, I think, a lot more flexible for consumers. I think over time, like in general, one of the things I think the pandemic really drove home for us was that resiliency is ultimately all about like how you have your own personal flexibility to be able to manage your costs. And also, I think, especially in a world in which technology is changing really rapidly, like Imagine a scenario in which your Tesla is like setting an over-the-air upgrade that's actually actively changing the car and making it safer. Your insurance rate's not changing, but that doesn't make any sense because you're not gonna get any— as many accidents anymore.

Right.

So I think insurance will largely start to be a lot more adaptive. And so you can think of this, you know, we think of this for a metric mile as being on a per-mile basis, but over time, I'm sure what we'll see is more and more opportunity for us to offer that in a per-mile basis differentiated based on whether the car is using different safety features, or autonomous, etc. And I think you'll probably start to see that in other lines of business as well. As insurers look at moments of time, as opposed to like a year-long policy.

Overall history. Yeah, yeah, that makes sense. I think related to, right, is like, as things are coming back. We were just chatting about events before we started the session, right? I— what— how do you see people's like driving habits now? Like, are— I know, like, I'm out in LA, I feel like people are still driving, like there's still traffic, so like the traffic never went anywhere. But I know in other places that's not the case. So how do you— how do you think about price per mile, right, still being relevant in times like this where it does seem to —how much people are physically driving? I know you answered a bit about automation, but how do you think about just in terms of people physically driving in their cars?

Yeah, so it's been really interesting. Maybe I'll start with just generally speaking, even before the pandemic, 65% of people drove less than the average. So the idea is basically there is a wide distribution. There's a lot of different kinds of behavior. And because of that, there are large groups of people that are overpricing for their— overpaying for their insurance. So that was true before and after the pandemic. But one of the things that we're seeing in the data today, which is really interesting, is number of miles driven in aggregate has largely returned to normal, but the times of day and the way the car is being used is very different. So we see instead of, you know, there being these very large spikes at commuting hours, it being actually a lot flatter throughout the day and seeing a lot more driving middle of the day or later in the afternoon. And I think part of the reason for this is, um, there is this somewhat, somewhat of like a bifurcation in behavior where there's a large group of customers now that are exclusively now working from home, or working from home for a large portion of their job, and others who are going back to their daily commutes. And because of that, now instead of having this kind of normal distribution of people who are driving more and less as we had before the pandemic, we see entirely different types of behavior, which means it kind of reinforces this need for more flexibility in insurance products.

Yeah, speaking of flexibility, is it something you see down the road as a trend of like almost surge pricing, right? Because if it is flexible and you see that most people are driving in between, you know, maybe everyone goes back to commuting times, it's 9 to 5, and if you're driving at 6 PM, PM, maybe it's cheaper, you know, for those miles that you're driving off hours. Like, do you guys ever think about building in something like that?

It's a really interesting idea. I think in theory it would be the mathematically correct thing to do, right? To like charge more during rush hour, charge less outside of that, because that's where most accidents happen. But I do think from like a customer standpoint, like I think putting myself in the shoes of being a customer now for however long, I don't think I want to have to constantly think about my per mile rate changing every minute or hour. Right. Um, but I do think there's opportunity to be able to, um, reward better behavior over time. Right. Sure. Um, and be able to say like, okay, you're, you're a better driver. So we're lowering your per mile rate overall. But I think there's this balance between flexibility and some level of certainty that you are going to save as opposed to having kind of the, the constant reminder of, you know, you know, what is my rate going to be today, tomorrow, etc.

Yeah, yeah, no, that makes sense. Um, another question for the audience, maybe it's a good one to, um, do as a final question unless we get some more. Um, how did you acquire and grow your early customers, right? Because a lot of this we've chatted about being disruptive and sort of being at the forefront, being flexible. Like, how did you guys grow, um, your customer base, and where do you think— and do you have any, like, certain things that you guys did in the early days that really set you up for the you have now?

Yeah, I will say the early days was all about kind of combination of just listening to customers one-on-one and then thinking really thoughtfully about the product. Because this is a space where there's an enormous amount of advertising. So we're not going to, as a small company 6 years ago or whatever it was, we're not going to win on Facebook or whatever it is. So we started by, I think, standing on bike bridges in Portland selling insurance to people who were running to work in the rain. We sold a few policies that way. Where we started to see real scale though is when we started to build product experiences that were really unique and delightful. And so this idea of helping people like avoid parking tickets or diagnosing car issues, we early on when we first launched in California, we helped one of our customers find their stolen car because we actually had, because the device is in the car, so we actually recover like 95% of all stolen cars. The story ended up making it onto the local news and it was like, 15 of us like crowded around the TV watching this. And we started to see that like there were stories like that and these little, like these unique delightful experiences that got people talking about the product and allowed us to build some initial, like an organic scale. And then over time that allowed us to then scale into kind of more mature channels that had more and more scale over time.

Yeah, that makes it— one more question from actually our founder, Esther. Do you have any learnings from being a SPAC success story? Because I know we hear the word SPAC a lot these days. Any learnings you could share with the rest of the audience of how it worked for you? Does it work for real companies with real revenues in the way it's worked for you guys?

For sure. No, it's a great question. I think in many ways, it should be treated in a lot of ways like a traditional IPO. Maybe thoughtful about your constituents and investors in the same kind of way. The thing that I think we were fortunate to have was a really strong SPAC sponsor, like the group that founded initially, because they understood insurance deeply and allowed us to reach the right kind of investors to be able to talk about—

Yeah, you need a good partner. Yeah. Yeah. That's always good. In anything you do, it's a good reminder. Well, Dan, I know we're at the half-hour mark. Thank you so much for answering our questions, your presentations. I know this is your first time speaking at Soundstruck, so hopefully we'll see you IRL soon. It was great.

Thanks so much for having us.