Telematics Insurance with Robert Smithson, CEO of Just Insure

Robert Smithson, Just Insure


Stewart: Welcome to another edition of Insurance AUM Journal podcast. My name is Stewart Foley, I’m your host standing with you at the corner of insurance and asset management, but not today so much. We are standing in the middle of insurance with none other than Robert Smithson. Welcome, sir.

Robert: Nice to be here.

Stewart: We’re thrilled to have you. You are interesting in so many different ways because you’re in this magical world that we want to learn about called Insuretech. You’re running a startup, you own an insurance company, which was one of my lifelong goals. So, what is Just Insure, and why’d you start it?

Robert: Okay, that’s a great question. You may not be able to tell this from my accent, but I’m actually not American. I’m actually British, and I came to the U.S. three years ago with a sports tech company, went through all the process of getting auto insurance. At my sports tech company, what we did was we priced sports events. That is, we worked out probabilities of events happening. I don’t know if you’re a big sports fan or not, but let’s say you’re watching the Rams versus the Chargers, and by the way, I certainly hope my American sports knowledge is not failing me here and they are actually in the same sporting thing.

Stewart: We share that.

Robert: Okay, good. So, you want to bet on the Rams. I’m sure you’ve seen the movie Casino with Robert De Niro, and in that Robert De Niro makes the odds and he becomes a wealthy guy, he does the sports betting in Vegas, but the world has changed. In the old days people would bet before the game, and then watch the game. Nowadays, people like to bet on the game as it’s going. It’s moved to in running betting. As I was very aware of how we build the systems, the technology we use, price, live sporting events, which player was where, who was on the field, how quickly were people running, really sophisticated stuff to enable us to price sporting events correctly. And sports by the way, it’s a much smaller market than insurance.

Robert: So, when I came to get auto insurance, I thought presumably there’s somebody out there who’s using technology to find out how I’m actually driving all the time, who is using technology to see, am I stopping and stop signs? Do I use my indicator before I change lanes? Do I change my driving behavior according to conditions? All those things, and I was shocked when I saw how poor the telematics offerings were in the insurance industry. And so, when I was lucky enough to sell my sports betting company, after having a little time doing some video podcasts, I thought, how can I use my risk pricing technology and experience in a much bigger market, and what market is bigger than U.S. auto insurance? It’s a $280 billion market, and nobody is doing real proper, real time risk pricing.

Stewart: Okay. So, the risk pricing I get. You used a term in there that I’m not familiar with. What are telematics? By the way, what we’ve skipped over is how you and I haven’t met until this minute. I’m watching something called Coverager, this email that comes in. A friend of mine who said, “You got to look at this.” I love this thing. I would tell you if you’re listening to this, just sign up for Coverager. I don’t know who these people are, I don’t know what they do, but it’s really cool. Anyway, Robert was in there a featured and I called him up on LinkedIn and stalked him like a weirdo, and I said, “Hey man, will you come on and do this podcast?” And he was so kind to do it. I should have rolled that out there before now, I apologize in advance. So, the telematics, what on earth is it? Not familiar.

Robert: In the old days, you had brilliant actuary, and becoming an actuary is no simple process. They would basically use risk tables and they’d say, “Oh, somebody in this zip code of this age has matched with approximately with this frequency,” and basically they were really heavy numbers of people. The old joke is that actuaries are people who’ve found accounting too exciting.

Stewart: You got to be careful in this room, man. You never know.

Robert: The actuaries I know absolutely don’t match that life stereotype, but the point is that they were never able to see inside the car, so they used demographic factors as things that correlated with risky driving behavior. Now, some of those are pretty obvious and easy. 18 year olds are bad drivers, by and large. I should say inexperienced drivers, more than likely to have accidents. I think that’s a controversial statement. They didn’t need to see into the car to know that if an 18 year old was driving, it was more likely to have an accident than if there was a 50 year old.

Robert: Those demographic factors, they were proxies for risk. They were basically ways of trying to capture the riskiness of the driver without being able to see how they actually drive. Telematics is the idea that we can use technology, whether things plugged into your car or your smartphone to try and see how you actually drive. That means tracking where you drive, when you drive, how fast you drive, being able to see, for example, we know where all the traffic lights are. If we look at your GPS track, we can see, “Ah, Stewart didn’t change his speeds, he went through that traffic signal, traffic was probably green. Ah, we saw Stewart started slowing down 200 yards from the traffic signal, traffic signal was probably red.”

Robert: Then all these ones where there was either late breaking or late acceleration, we can use that to analyze that, and work out how far from the traffic signals you make the decision whether to accelerate through or to break. We can work out, we can guess based upon your actions when that traffic signal began to change. Our conjecture, and this isn’t proven yet, our conjecture is that people who take risks with going through as it’s turning to red are more likely to have accidents than people who do not. So, what we’re trying to do is to go beyond just using who Stewart is, and instead to say, “How does Stewart drive? What is his real risk of an accident?” By using real risk metrics, ie: where he drives, when he drives, how he drives, not just how old he is.

Stewart: That’s an amazing capability from my perspective, also kind of scares me because I speed constantly, but forget about that for a minute.

Robert: Well Stewart, shall I tell you some good news? Speeding does not correlate very well. Well, moderate speeding does not correlate very well with accidents. Where do you live, Stewart? What part of the world?

Stewart: Just outside of Chicago.

Robert: Just outside Chicago. I don’t know Chicago well at all, but let’s imagine it’s 4:00 PM, Sunday afternoon, the sun is shining, you’re on a freeway, there’s nobody else on the freeway, and you do 85 miles an hour. That’s 20 miles above the speed limit. That’s not dangerous behavior. The fact that the conditions are perfect, there’s nobody else around, it’s not dangerous to go above the speed limit. On the other hand, if it is pouring with rain, the road is busy, it’s 9:00 at night on a Friday night, there’s people coming home from bars and from the office, and the flow of traffic is going 50 miles an hour, if you’re weaving in and out of vehicles to get to 60, even though you’re not breaking the speed limit, what you’re doing is extremely dangerous.

Robert: One of the things that we really noticed was that people who change their driving and their speed according to the conditions are safe drivers, people who don’t change their driving according to the conditions are unsafe drivers. So, unsafe drivers might drive at 60 miles an hour irrespective of whether it’s snowing or sunny or busy or quiet, whereas a safe driver will properly speed where the condition is good, but will probably drive a lot slower if conditions are bad.

Stewart: Very interesting. That’s really cool, you obviously have an amazing amount of technology there. Are there groups of people that you are trying to work with that you’re focused on that benefit from this approach more than others?

Robert: Yes. It’s a great question because there have been other people that have tried to do telematics in auto insurance, and shout out to Roots, who’s based just outside Chicago. They’ve done a good job, I think we’re better, but they’ve done a good job. But Roots, and a few of the other players in this space, they all want customers who are people like you and me. People who are roughly well off, people who probably own their own home, and maybe have a couple of cars. They like well off customers because well off customers don’t turn very often. If you’ve got them on, you keep them.

Robert: Our view is that the ideal customer for a telematics insurance product is a lower income customer, and the reason for that is that many of the factors that auto insurance companies use, the set rates, while completely correct in aggregate, are often quite discriminatory towards low income consumers. For example, zip code matters. If you live in a poor part of town, you’ll be paying more for your auto insurance, even though the biggest component of your auto insurance is really your risk of having an accident and paying out somebody else. You will also find that auto insurance companies use credit scores.

Robert: Now, they use credit scores for a good reason, which is that if you are a 28 year old man earning $45,000 a year and you have a shitty credit rating, that’s because you have an attitude towards risk, you like it, and that correlates with driving less faithfully. On the other hand, if you’re 48 years old, earning $25,000 a year, you don’t have a bad credit rating because you are risking, you have a bad credit rating because you don’t have any money. Our offering is designed to basically find those people that have been discriminated against, or it will be typically attracted to people that will be discriminated against because they come from these things. The savings they get are potentially enormous.

Stewart: So, we’ve talked a little bit about the technology and the insured, what about the roads themselves? The amount of data that you have, and that you’ll continue to collect, and you’re a new company and we can talk about it, I’d love to talk more about the history and how you found that. What are the implications for the roads themselves, or is there traffic flow design issues, or is there any interface there?

Robert: It’s an interesting question because traditional auto insurance doesn’t really encourage you to drive more safely. If you drive at 90, and unless you have an accident, your insurance rate is not going up. With us, we are working on giving our drivers real time feedback to help them become safer drivers. From their perspective it saves them money if they follow our instructions, and also if they don’t follow our instructions, it tells us something about them as well, which is also interesting. Both myself and my co-founder, Murray Macdonald, we’ve both lost friends and family members in auto accidents. From the start, we’ve thought that insurance here can be a force of good, a force of change. It can be something that saves lives because if it can make people better, safer drivers, fewer people will die.

Stewart: Is your technology different than other… Whenever I see auto apps for other insurance companies, I assume that they’re capturing data on how I drive, and I’m distrusting of that. Just being honest. How is your technology different or is it different?

Robert: The actual data collection elements, ie: using your phone or an OBD2 dongle to track your GPS, your acceleration, and various other things, that’s very similar. The other guys, and when I say the other guys, I mean mostly I’m talking about Progressive, I’m talking about the other big auto insurers, all of whom have telematics based products. The difference between us and them is they have all pretty much chosen to buy in technology. They’ve gone to, with the exception of Progressive, they’ve gone to this Italian company called Octo, and they licensed Octo’s technology.

Robert: Why is that? It’s because they’re not technology companies. The other thing is, if you look at how all of the other big insurers work, they say, “Stewart, you want to sign up for our telematics product, go through the same way, answer age, level of education, if you own or rent your house, are you married, do you have children?” Et cetera, et cetera, et cetera, et cetera. You know the 40 questions. And then they say, “Okay Stewart, your price is $140 a month, and if you drive really well, you could earn a 30% discount.”

Robert: That’s the pitch. It’s all the same long signup process, and in some vague way, you could earn a 30% discount. There’s two problems with that. Firstly, we’ve tracked 13,000 peoples’ driving habits. We’ve paid for last year to drive around. There’s not a 30% difference between the accident likelihood of a good and a bad driver, there’s like a 3000% difference. So, they’re only capturing a very narrow range of risk outcomes.

Stewart: That’s interesting. That is really interesting.

Robert: That of course means that they can buy into it because they don’t have to be too exact because they’re only getting a really rough thing. We on the other hand, if you drive really well with us we could cut your price to as low as $0.02 a mile, and that’s the cheapest theoretical driver… By the way, I’m not sure whether somebody actually could exist getting $0.02 a mile], but they certainly could get three. If somebody was literally, they only drive at the safe times of day, they only drive on roads that we know don’t have many accidents, and they demonstrate good driving behavior, they could easily be paying just $0.03 a mile.

Robert: On the other hand, if you drive back from bars to home at 2:00 in the morning on Friday night regularly, and if you drive along streets where there are lots of accidents, and if you’re not very good at stopping at stop signs or indeed traffic signals, then you could well be paying $0.45 or $0.50 with us because that’s how big the risk is. Not 30%, it’s 13 times, and we needed to have a technology that was good enough, to be able to deal with the kind of resolution needed, to be able to pick up all those outcomes.

Stewart: That’s very cool. How big is the market for telematic insurance?

Robert: There is going to be a significant portion of the country that won’t want it. You don’t want to be tracked. There’s also, the reality is, if you are a well off consumer living in an excellent town, you’re probably not going to save that much. In which case, why go to the trouble? If you are a person whose income is variable and who only drives when they’re working, and is being charged per mile, that’s incredibly attractive especially if we also reward people better driving. Right now, 13% of cars in the U.S don’t have legal minimum insurance. That’s our starting market. We want  those people because they can’t afford it because what’s being produced right now is too expensive for them. We are perfect for those customers. We’re also perfect for anybody who doesn’t drive that much or anybody who is significantly safer than their demographic or those have infrequent incomes. I think over time this idea, and I’m not saying we’re going to capture all of it, but I think this idea could easily be a quarter to a fifth of the U.S. auto insurance market, and we should be the leading player.

Stewart: That’s a big, big market too.

Robert: Worth getting out of bed for.

Stewart: Absolutely. So, on the wrap up into things here, I always tend to want to get to know our guests on a little different level. My trademark or whatever you want to call it, frequent, worn out question comes from the fact that I was a college professor for a long time, and I saw and see a lot of students come up through college and then graduate and move out into the world. We always talk about the opportunities that exist in the insurance market in all sorts of different opportunities, and they often are not familiar with the insurance industry or have a bad impression of the insurance industry or whatever it may be. So, my final question for you is what would you tell your 21 year old self?

Robert: What would I tell my 21 year old self? Stop smoking. I think that would be number one. What else? Exercise more, be very good at risk pricing. Actually, I’ll tell you what. The sole most important that I’ve learned in life and in business is that the biggest cost is opportunity cost. If I do this, what is it that I’m not doing? What is it that I am missing out on because I’m doing it this way, and that’s a recipe for… Remember that because being in an unhappy relationship or being in an unhappy marriage is opportunity cost. Being in the wrong job, being something you don’t enjoy is opportunity cost. And in business, remember to keep your eyes on the important thing because what you’re doing might be valuable, but it may carry big opportunities. So, that is my sole bit of advice.

Stewart: That is great advice. We really appreciate that. I think that whenever you have very successful executives and entrepreneurs, and I think students and people early in their careers benefit tremendously from things like that, and that’s not something that anybody’s going to learn in a textbook. I know everybody really appreciates that. So, thanks for taking the time to spend with us. We appreciate it. I learned a lot, man. It was very interesting.

Robert: I enjoyed it too. Thank you so much for inviting me on, and I look forward to seeing it and sharing on my channels.

Stewart: Well, thanks so much. Thanks for listening. Like us, download us, tell your friends. This is the Insurance AUM Journal podcast.

Robert Smithson is the founder and CEO of Just Auto Insurance, a telematics-based
auto insurance company. Just was founded in January 2019 when Robert realized that the
consumers who would benefit most from a telematics-based product were those on lower
incomes, who pay too much for auto insurance because of the use of measures such as credit
rating and zip code in setting rates. Just launched in Arizona at the end of March 2020 and is
growing rapidly. Before founding Just, Robert was a partner at a fund management firm and
founded several businesses including Genius Sports (2000) and PythonAnywhere (2012).