Rewriting the rules of auto insurance with Robert Smithson

When Robert Smithson moved from London to Los Angeles in 2017, he brought with him a passion for building tech companies and a strong belief that U.S. auto insurance was ready for a major change. A philosophy graduate from Cambridge with a background in finance, he had already founded and sold Genius Sports and PythonAnywhere. He saw a huge opportunity to completely rethink how risk was priced in the insurance world.

Today, he leads Just Insure Inc., which boasts the lowest loss ratio among auto insurance carriers in the U.S., according to Smithson. Instead of locking drivers into year-long contracts, Just offers 90-day policies and adjusts prices based on real-world driving behavior. The company uses smartphone data to understand not just how people drive, but also when, where, and under what circumstances. This innovative approach has allowed Just Insure to keep loss ratios more than 20 points below the industry average, all while offering customers savings of up to 80%.

* * *

We recently sat down with Robert to talk about his journey from Cambridge to California, the thinking behind Just Insure’s tech-first strategy, and his vision for making auto insurance both fairer and more profitable.

* * *

You studied philosophy at Cambridge. How has that influenced your approach to building companies?

Philosophy taught me a way of thinking that’s both analytical and abstract. You learn to question assumptions and build arguments from the ground up. In business, that’s incredibly valuable. I tend to tackle problems by breaking them down to their core logic before deciding on the next steps. It’s also given me a healthy dose of skepticism about traditional wisdom, whether it’s about how insurance should be priced or how startups should raise capital. Philosophy really pushes you to think critically, which is essential when you’re building companies that aim to disrupt established industries.

You founded two successful tech companies before Just Insure. What’s the common thread connecting Genius Sports, PythonAnywhere, and now Just?

Data. The common thread in everything I’ve done is using information to make better, fairer decisions. At Genius Sports, we used data to understand and predict sports performance and betting outcomes. Now at Just Insure, we use data to understand and price driving behavior. I’m a huge believer that there’s so much information out there; we just need to harness it and analyze it correctly. This really began during my time in finance at Goldman Sachs and other firms, where I was always searching for that informational edge. Looking back at the very early days of Genius Sports, we were collecting prices from sportsbooks, and I was constantly asking myself: what can we do with this data? Ultimately, it’s all about replacing subjective judgment with objective evidence.

What drew your attention specifically to the insurance industry?

Insurance is fascinating because, much like sports betting and finance, it has two types of players: those with incredible intuitive judgment for pricing risk, and those like me who are all about numbers and systems. Insurance also has what you could argue is the first incarnation of modern data science: actuarial science. However, I think that made the industry slow to realize that change was on the horizon. Traditional actuarial science relies on slowly changing demographic factors like age, zip code, sex, and marital status. This approach missed the enormous amount of data now being collected in real-time about drivers. We could suddenly measure actual driving behaviors instead of just proxies. That’s when I knew there was an opportunity to make insurance fairer and smarter. That’s how Just Insure was born: a technology company first, an insurer second.

You came to the U.S. in 2017. What made you decide to build this company in America rather than the U.K.?

I came to the U.S. because my previous company, Genius Sports, was focused on pricing risk in the sports betting market, and online sports betting was about to become legal here. Once I was in the States, I started looking at other industries where real-time risk pricing could make a significant difference. Insurance immediately stood out. It’s a massive, $300 billion market that’s fragmented, and the way risk was being priced seemed outdated compared to what I’d seen in other sectors. The U.S. startup culture is also more aggressive; people move faster and are willing to take bigger risks. I liked the idea of combining that boldness with the analytical rigor of the British approach.

For readers who aren’t familiar, could you explain telematics? What exactly are you measuring?

Telematics is essentially using technology to measure how, when, and where people drive. We collect data from drivers’ smartphones and vehicles and feed it into our AI pricing models. This allows us to set premiums based on actual behavior rather than just demographics. In simple terms, safer drivers pay less because we can truly understand how safely they drive. But it’s not just about acceleration and braking; we’ve analyzed over 10,000 behavioral and contextual signals. This includes the time of day, whether the driver is distracted by their phone, and whether they’re on a freeway at night or a side street during daylight. All of this information contributes to our proprietary model.

How is your approach different from traditional insurers?

Traditional insurers tend to look backward, pricing based on averages rather than individual behavior. Our technology looks forward, constantly learning and adjusting based on real-world driving. We built our own policy management system and pricing model from scratch, instead of trying to layer telematics on top of old, legacy systems. That’s why our loss ratios are more than 20 points better than the industry average: we identify and price risk dynamically. We also do something unique: we issue 90-day policies and reprice every 90 days. If a driver’s behavior becomes too risky, we raise their price, and if they leave, that’s actually a good thing for us. It helps us maintain a healthy risk pool.

Other insurers like Progressive and State Farm also use telematics. What makes Just Insure stand out?

Traditional insurers rely on long-term customer retention, so they issue six- or twelve-month policies and try to hold onto every customer. Our model works differently. Competitors often focus only on the car, using dongles or odometer readings. We, however, focus on the person, capturing their behavior, context, and risk appetite. This holistic approach, combined with our rapid churn management, is why our loss ratios are 20 to 30 points better than the industry average. We’re constantly refining how we assess risk rather than waiting for an annual renewal cycle. It’s a live feedback loop between behavior, data, and pricing.

What role does AI play in your operations?

AI is at the very core of how we operate. It helps us interpret telematics data within its context, allowing us to distinguish risky driving from safe, defensive maneuvers. For example, a sharp brake might look risky, but if it’s to avoid an accident, it’s actually evidence of good driving. We built our models at Just Insure with that in mind. AI also helps us predict claim likelihoods with a level of precision that traditional actuaries simply can’t match. Our Head of Data Science, who previously led pricing at Uber, designed this model with the same rigor you’d expect in dynamic pricing environments. Ultimately, AI makes our insurance fairer, faster, and more efficient, while also giving drivers immediate feedback on how they can save money.

How do you address concerns about data privacy?

Data privacy is absolutely fundamental to what we do. We are completely transparent with our customers about what data we collect and how it’s used, and we never sell or share it without their consent. We also anonymize and encrypt driving data so it can’t be traced back to individuals. If you want people to trust you with their data, you have to earn that trust every single day. When customers share their driving habits with us, we use that information to price them in the fairest way possible.

You’ve mentioned that traditional auto insurance often penalizes lower-income Americans. Could you elaborate on that?

Traditional auto insurance can be particularly expensive for lower-income Americans. They often live in zip codes that have higher auto insurance premiums, and they’re more likely to have lower credit scores. This means that auto insurance isn’t just more expensive relative to their income than for more affluent individuals, it can actually be more expensive in absolute terms. If you look at the data, people with the worst credit scores can pay more than three times what those with excellent scores pay. It’s incredibly difficult to break out of poverty and improve your credit when your insurance takes up an ever-larger share of your income. And this often leads to too many Americans driving without insurance: typically about 14%, compared to less than 1% in Europe or Canada.

How does Just Insure tackle this problem?

When we started Just, we saw this as an extraordinary opportunity. We asked ourselves: how can we create an auto insurance product that truly works for lower-income Americans? How can we use real behaviors to encourage safe driving, increase the number of insured people on the roads, and reduce the number of accidents? With a traditional insurance product, you don’t really personally benefit from driving more safely. With telematics, that all changes: if you exhibit safer behaviors or don’t drive as much, you can personally benefit. We have customers who pay 80% less than with traditional insurers. In fact, we’re so confident in our telematics model’s ability to price risk that once customers have 3,000 miles of driving data, assuming they’ve abided by our terms and conditions, we completely remove the impact of credit on their price.

Credit scores and zip codes aren’t inherently discriminatory. Why is it unfair to use them?

While credit scores and zip codes aren’t inherently bad ways to measure risk in general, they often lead to systematic biases against individuals. A good driver in a low-income area might pay double what a worse driver pays in a wealthy zip code, simply due to statistical correlations. That’s just not fair, and technology now allows us to do better. We price people based on how they actually drive, not on where they live, their education, or how much they earn. Telematics makes insurance more equitable by turning it into something people can control. You can’t change your credit score or zip code overnight, but you can change your driving habits. That gives customers agency, they’re not just passive buyers, they’re active participants in how their risk is priced.

What have you learned about when and where accidents are most likely to occur?

Nighttime driving, especially between midnight and 3 a.m., carries a much higher risk. Distracted driving on highways is also a major factor, while short, local trips during daylight hours tend to be safer. These thousands of inputs help us identify not just obvious risks, but also subtle patterns that wouldn’t show up in simpler models. The biggest savings go to careful, low-mileage drivers. But we also reward improvement: as your driving behavior gets better, your premiums drop. That positive feedback loop encourages safer roads overall.

What did your experience in finance teach you about entrepreneurship?

Finance taught me discipline. You learn how to evaluate risk, manage capital efficiently, and make decisions under uncertainty. That experience has been invaluable as an entrepreneur, it’s one thing to dream big, it’s another to make sure the numbers work. My time in fund management trained me to think probabilistically, which is essentially what good business is all about. Every business I’ve built starts with the same principle: understand the problem better than anyone else. Whether it’s finance, sports, or insurance, the pattern is the same: find inefficiencies, apply data and technology, and solve them elegantly.

Gary Tolman recently joined as Chairman. How did that collaboration come about?

The collaboration happened quite naturally. Gary and I had been in touch through mutual industry contacts and quickly realized we shared the same philosophy: that data-driven insurance can truly improve people’s lives. Gary’s arrival brings deep insurance experience and a steady hand. He’s been through the entire journey of scaling digital insurers before, from early growth to acquisition at companies like Esurance and Noblr. That perspective is invaluable. It allows me to stay focused on innovation while knowing that governance and structure are in great hands. I bring a technology-first mindset and experience scaling startups quickly. Together, we have a great balance of old-school discipline and new-school innovation.

What’s your advice for other entrepreneurs looking to enter insurtech?

Understand the regulatory side before you even start. Insurtech is complex, and you can’t just ‘move fast and break things.’ But don’t be intimidated, the industry is full of opportunity if you can navigate the rules smartly. Build something that genuinely improves outcomes for customers, not just something that looks good in a pitch deck. In my case, working through state-by-state regulation was a challenge, especially as a newcomer to the U.S. But those hurdles also meant that if we could succeed here, we’d build something very defensible.

You started in Arizona and Nevada. Why those states, and what’s your expansion strategy?

Both states offered straightforward regulation and relatively low catastrophe risk. With less exposure to hurricanes or major weather events, they were safer environments to test our model. Arizona gave us a diverse driving population to start with, and Nevada was a natural next step geographically. We’re now filing in multiple states simultaneously and expect to be in seven states by the end of 2026. Texas is a major priority for us.

How have you managed to achieve growth while staying profitable?

Our growth has come from focusing on the fundamentals: low loss ratios, efficient capital use, and high customer satisfaction. We’re not chasing vanity metrics; we’re building sustainable unit economics. Our technology allows us to quickly churn unprofitable customers and retain high-margin ones, which compounds over time. We’re positioning ourselves as the most profitable auto insurer in America, not necessarily the biggest, but the smartest. By pricing risk more accurately than anyone else, we can offer lower premiums to good drivers and still make more per policy. In a $300 billion market, accuracy definitely beats size.

What are your main milestones for the next year?

Continued expansion into new states, scaling our Just Unlimited product, and reaching sustained EBITDA profitability in early 2026. We’re also deepening our agency partnerships, which are already showing fantastic traction. Each of these steps builds upon the next stage of our growth.

Cars are evolving rapidly with electric, connected, and autonomous features. How do you see insurance keeping pace?

Insurance absolutely has to evolve alongside connected and autonomous vehicles. More data will become available, and pricing will become even more usage-based and dynamic. That’s exactly the direction we’re already heading in. I believe usage-based and on-demand insurance will eventually replace traditional policies. Consumers increasingly expect flexibility and fairness. It won’t happen overnight, but the trend is very clear.

What’s your long-term vision for Just Insure?

My long-term vision is a national insurer with millions of customers, strong profitability, and a reputation for being the fairest and most innovative option in the market. Insurance is one of the biggest markets in the world and arguably one of the least efficient. When I realized how broken the incentives were, how much opportunity there is to make a real difference, and how big the potential market is, I couldn’t resist. It’s also an industry that directly affects people’s financial lives. If you can use technology to make it cheaper and fairer, you’re not just building a business, you’re giving people the right incentives to make smarter, safer driving decisions.

What’s the biggest lesson you’ve learned as an entrepreneur?

The biggest lesson is that risk and opportunity are two sides of the same coin. The key isn’t to avoid risk; it’s to understand it better than everyone else. That’s true in life and in insurance. If you know what you’re taking on and can measure it properly, risk actually becomes a competitive advantage.

If you could give your 21-year-old self one piece of advice, what would it be?

Be patient but persistent. You don’t need to have it all figured out immediately, just keep learning and taking calculated bets. Most people overestimate what they can do in a year and underestimate what they can do in ten.

About Just Insure

Just Insure  is a technology-first auto insurance company that uses telematics and AI to price policies based on actual driving behavior rather than demographics. Founded in 2019 and headquartered in Los Angeles, the company currently operates in Arizona and Nevada, with plans to expand to seven states by the end of 2026. For more information, visit https://just.insure