Climate change is challenging the limits of traditional insurance by creating new risks and exacerbating old ones at an unprecedented scale. As weather events increase in frequency and severity, many businesses find themselves underinsured or unable to secure coverage. Arbol addresses these growing protection gaps with parametric insurance, which quickly assesses damage based on a predetermined set of markers informed by historical data. In an interview with McKinsey’s Fabian Metzeler, Siddhartha Jha, Arbol’s co-founder and CEO, discusses how parametric insurance creates extra value in the insurance industry, Arbol’s obsessive focus on cashflow over growth, why AI has been essential to streamlining Arbol’s sales process and operations, and what he wishes he’d done differently while fundraising for the company.
Fabian Metzeler: For the uninitiated, could you explain the difference between parametric insurance and the more traditional insurance products we’re familiar with? Why is parametric insurance specifically gaining momentum in areas that are highly exposed to weather-related risks?
Siddhartha Jha: Parametric insurance is starting to gain traction as people become more aware of it. The primary difference is that parametric insurance is triggered by a data point rather than a subjective loss assessment. In a traditional insurance contract, an adjuster or specialist comes to your property after a natural disaster like a flood or hurricane, thoroughly documents all losses, and determines a loss amount. This subjectivity can introduce a number of inefficiencies, delays, disputes, and sometimes even fraud into the claims process. It also makes scaling coverage a challenge. More parts of the world and types of insurance have reduced coverage or are experiencing an increase in delayed payments. In general, people are unsure when and how much they’ll get paid for their claim.
Parametric insurance tries to address those challenges by using increasingly granular data sets to help with this process, especially when traditional insurance lacks sufficient coverage or has many exclusions. In this way, climate insurance is an ideal parametric use case because we continue to collect more precise data for the weather and climate.
Fabian Metzeler: Could you give an illustrative example as to how parametric insurance can be used in an industry such as agriculture or energy?
Siddhartha Jha: An example in agriculture might be when a farmer is worried about deficient rainfall over a crucial growing period for their crop. Parametric insurance would use a highly granular rainfall measurement that applies to their farm area. If rainfall is 50 percent below a long-term average, the farmer will get a payout. For that kind of coverage, farmers will pay a premium. This also could apply to a processor that makes flour. They might be exposed to a drought in the areas where they source the wheat from. Processors aren’t growing the crop directly, so it’s difficult to get insurance for their supply chain, especially if it includes thousands of farms. There wouldn’t be enough adjusters to cover the area.
A parametric insurance contract, on the other hand, could easily cover a district or a number of districts. It starts to add value in situations where sending an adjuster would be too difficult. This might apply to, for instance, a utility that is worried about a summer heat wave. When temperatures spike, everyone starts running air conditioning. Utilities are obligated to supply power, but they may be limited to a certain amount for that period because their plants only produce so much. If they have to deliver more power unexpectedly, they can’t cut power because that will be costly in other ways. Instead, they have to procure more power in the live spot market, where they might pay ten, 20, or possibly a hundred times more than average. To mitigate this situation, utilities often buy temperature contracts to protect themselves. That way if there’s an unexpected heat wave, they get a payout that helps cover some of the cost of procuring spot power.
Fabian Metzeler: How did you come up with the idea to found a new company in this highly competitive space crowded with big, competent players with huge financial reserves?
Siddhartha Jha: There was a confluence of factors. One was that I was an outsider who didn’t overthink the problem. I was from the commodities world where weather and climate are big problems. So we started with a blank slate when it came to thinking about climate impacts in an insurance context. The makeup of our team was crucial to tackling this problem. We assembled a team that had expertise in new technologies, financial markets, and regulatory and legal aspects, to name a few.
Our founding team’s position wasn’t to beat the incumbents but rather to figure out where there were gaps in coverage. We asked what sort of products a farmer or a utility would need besides traditional insurance. Then we brought new players into the market who could service those needs, because those gaps weren’t being filled by incumbents due to the risk, the impact on their credit ratings, and other issues. By identifying a narrow problem, solving it, then showing that we could execute, we started to create a niche.
Fabian Metzeler: The heyday for start-ups is, at least temporarily, over. How did you and Arbol navigate the sudden paradigm shift from “growth at all costs” to “better become profitable ASAP”? In which areas did you need to pivot to adopt to this change of direction?
Siddhartha Jha: My co-founder had been in the markets since the late 1980s, while I spent the first five years or so of my career in interest rates and macroeconomics. We approached fundraising with a different viewpoint from someone who has a purely tech start-up background. We never followed the growth-at-all-costs model and were therefore out of sync with the fundraising market during those years. I remember that our seed round deck talked about when we would achieve positive cash flow because we knew that the interest rate cycle had to change at some point and funding would dry up, causing a big problem. Our focus was confusing to prospective investors. We obsessively focused on how every dollar was spent and how much revenue it could generate. And when the markets changed, we were in a much better spot compared with many other firms that had overly emphasized the growth aspect.
Fabian Metzeler: How did you navigate the transition from “scrappy” to “professional” as a fast-growth start-up? That is, how did you time and sequence the maturation of organization while retaining the culture and agility, especially in an industry with a limited appetite for experimentation?
Siddhartha Jha: One of the things we’ve focused on is hiring a mix of entrepreneurial and deeply experienced people who can take on different aspects of our roles. That’s easier said than done. Generally, as people gain more experience, they tend to become less entrepreneurial and more entrenched in their roles. To achieve some balance, we try to have someone start off in a deputy role to learn the ropes. Then they can start taking on more responsibilities. We’re still fostering this independence in areas like sales efforts, where we want to expand past founder-led strategic partnerships into having people who build partnerships on their own. It’s difficult to train people in business development because it includes educating customers on the product, embedding it into their systems, and checking all the legal and regulatory boxes. But that’s where building a cohesive team who can work with each other and think outside the box is so crucial.
Learn to Leap interviews
Fabian Metzeler: How did AI help you work with the large and complex sets of data parametric insurance is built upon?
Siddhartha Jha: We’ve used AI since the start of building the company. My own background is in graduate-level statistics, and we have a deep bench of expertise in AI. We developed our underwriting, pricing, and risk management to be an AI engine that has been learning since day one. This was not a “nice to have” element of our process; it was a necessity, given that climate dynamics are incredibly complex and produce vast amounts of data. Every day, we have tens if not hundreds of millions of data points that have to be processed and understood. And within the system, a climate event happening in East Asia can affect another one happening in West Africa. It is nearly impossible for a human being to keep track of that. Actually, I would say it’s definitely impossible.
Fabian Metzeler: When I look around Arbol, I count about 50 employees or so. How do you use technology to ensure you can grow in the leanest way possible? And how has the strong momentum of generative AI and AI created new opportunities for you to streamline your organization?
Siddhartha Jha: We’ve found technology to be especially useful in streamlining the sales process. While your AI engine will give you a price for your product, we found that there’s too much human intervention in the structuring of products in the sales process. Streamlining the sales process looks like receiving pertinent information from a client and then deploying bots to run different combinations of structures. Those structures can then meet various targets for the client, such as their premium budget, and link that information with the pricing engine. For us, all these actions created an efficient sales process that we used to grow revenues from $0 to $170 million gross with just two salespeople.1 How? Because they had an incredible array of tools at their fingertips to deploy a structure.
Our next area of focus has been the back-end systems and reducing human touches on the operations process. That part doesn’t need AI but rather automation. We’ve used automation to streamline aspects of the sales process from our early days, like shifting from blockchain white paper to smart contracts. We’ve built our technology stack incrementally with an eye on demonstrating its use in generating revenue or reducing costs. With a “build it and they will come” approach, you’ll end up with an overly engineered platform that distracts from the reason you’re in the business—to sell a product.
Fabian Metzeler: How would you describe your personal management and leadership style? How do you enable your broader, more recent leadership team to replicate essential aspects of your leadership that are important for Arbol’s success?
Siddhartha Jha: The overall ethos I’ve tried to impart has been that we have the leadership team that is responsible for management, and below them, we have pods that are independently responsible for their deliverables. That could include a legal pod or an accounting pod but also a sales pod where you have someone who faces the client and someone who faces the risk capacity. Each pod has specific deliverables. My role is to ensure that the heads of each pod understand what the priorities are and how they can cooperate with each other. For example, if the sales pod needs regulatory approval, they’ll need help from the legal pod, and so on. Essentially each pod is responsible for managing what equates to their own mini companies.
Something that’s been really helpful in navigating this vision is making a consistent effort to talk to all our employees, including some of the junior people. This gives us an inside perspective on the inner workings of the company that you don’t always get from the top brass. If they are not clear on the priorities, then they can’t achieve their deliverables or understand their impact. Having one-on-one discussions with people from all over the firm helps make them aware of the overarching goal.
Our leadership team is very collegial. We have always made decisions in a group setting. At times, we’ll have disagreements, but we work hard to get on the same page. We recognize that everyone here is not just smart but also brings diverse and deep experience to the table. A big part of my work is to distill those experiences, not fight them, and understand what information or expertise I’m missing from my perspective.
Fabian Metzeler: In hindsight, what are some pieces of advice on what to do differently that you would give to your younger self?
Siddhartha Jha: There are so many pieces [of advice] that it’s hard to pick. One is that I would put more time and effort into hiring. When the firm was young and growing fast, we were just trying to fill roles, so we took some shortcuts in that aspect. I’ve learned that getting the right people into the right roles doesn’t necessarily mean hiring externally. Getting that right has been an important job for management to learn over time.
Another thing I would do differently is to be a lot of more organized about my fundraising. I could have done a better job getting to know prospective investors before thinking about fundraising. Because our growth was so fast, our attention was scattered. But even when fundraising doesn’t seem imminent, it’s important to get to know investors early so that they are comfortable with you and you can assess their expectations for their investments. You can also gauge what they might be like as a partner. Investor partners have the potential to give you valuable advice on specific problems you might have in your business or your overall points of vulnerability. The earlier you get to know your investors, they earlier they can see that you can execute, which lays a strong foundation for your partnership later on.
