It’s hard to believe I’ve already been at Coalition for one year — what a ride it has been. At Coalition, we are hard at work solving cyber risk. Our approach is unique; we package cyber insurance with cybersecurity tools and services to protect our customers from digital risk. The result is complete alignment between Coalition and our policyholders. We not only write the insurance policy, but also take proactive measures with the policyholder to reduce (and hopefully eliminate) the risks they face as part of operating in a highly connected digital world. For this update, I thought it would be fun to provide an insider’s view of insurtech after immersing myself for a year at Coalition.
Coalition is an insurance company built with technology at its core, hence an ‘insurtech’. Before engaging with Coalition, I had never even heard of insurtechs. In hindsight, it probably should have been obvious that one of the largest industries in the world would be going through a technology transition of epic proportions.
Here are a few data points about the insurtech market. This McKinsey report from 2020 sizes the insurance market as $5 trillion globally with $1.6 trillion in the property and casualty segment where Coalition plays. Next, the Willis Towers Watson Q2 2021 quarterly insurtech briefing highlights investment activity in the space. YTD 2021, there has been $7.4B in investment across 308 transactions. That’s already exceeded $7.1B in total investment for all of 2020. What makes insurtech so attractive is the chance to modernize a multi-trillion-dollar industry. In everything from risk selection and pricing to digital distribution to analytics to new services and solutions, there is a massive opportunity to bring insurance forward into the modern age.
Coalition’s role as MGA
As you might expect, the definition of an insurtech is broad. Coalition is a Managing General Agent (MGA), a specialized type of insurance provider that has underwriting authority from an insurer. We perform certain functions ordinarily handled only by insurers, such as binding coverage (binding = purchasing in insurance speak), underwriting and pricing, and settling claims. We have talked publicly about the relationships we have in place for cyber insurance with carriers such as Swiss Re and Arch.
When Coalition writes a policy, it is a Coalition policy, but one or more insurance carriers take the actual financial risk. This means, Swiss Re and Arch have confidence in our data-driven underwriting practices and the way we assess a potential policyholder's risk posture.
When the product is not the product
At Coalition, our technology organization is not building the actual insurance product we are selling. As an MGA, Coalition’s product is an insurance policy. Our CEO, Joshua Motta even likes to joke that we are in the business of selling PDFs. Having said that, I haven’t found that the distinction between the technology products (quoting platform, underwriting engine, etc.) and the financial product (the insurance policy) really drives much difference for us. I liken it to a second layer of product management. We have a traditional product management function that is driving the evolution of our product capabilities around quoting, distribution, policy management, and underwriting. We partner with our insurance team who drive policy and coverage changes, underwriting guidelines, or securing reinsurance contracts and relationships.
Insurance is inherently a complicated product and the bigger implication from a technology standpoint is what we like to call ‘insurance-in-a-box.’ For any insurance product like cyber, we need to model coverages, sub-coverages, limits, retentions, and more. Then, each product may be customized for different carriers or geographies. insurance-in-a-box is the notion that we need a flexible, configurable engine that makes it easy to add new products to our distribution and underwriting platforms as we scale.
One of the bigger surprises for me at Coalition was that 40% of our submissions were manual. These submissions originate from one of our customer support channels, typically chat or email. This is a sign of just how much of the industry still operates using legacy processes and tools. It also signals significant opportunity for future innovation. Those manual submissions are a disparate variety of different broker-specific forms. Over time, we need to provide a frictionless quoting experience to serve brokers however they work, whether a broker is using our first-party application, a technology partner that has integrated APIs, or a hand-written obscure form. Getting to a future where all submissions are automated will take substantial investment in technology around text extraction and processing including optical character recognition (OCR) and natural language processing (NLP).
Underwriting at scale
The scope of underwriting is deceptively simple — decide whether to decline or accept a quote and assign a price.
The technical challenge inherent in underwriting at scale, however, is fascinating. It is fundamentally an optimization problem.
We can take larger risks that yield more premium but may risk an undesirable loss ratio. Similarly, there are times when we trade accuracy for faster underwriting. Underwriting effectively at scale is one of the hardest things for insurtechs to get right, and it is our biggest differentiator as one of the leading cyber insurance providers.
Underwriting starts with data. A small amount of that comes in from the broker on the application. Coalition invests heavily in world-class data collection built around scanning the internet. We walk the entire internet, scanning for open ports and associating technologies we discover to known vulnerabilities. Additionally, we consume several different open and licensed data sources. We then combine these terabytes of information with our own claims data and other sources to power machine learning in our risk evaluation process. This gives us a complete picture of the potential insured’s digital enterprise and their associated risks when deciding whether or not to offer a cyber insurance policy, and at what price.
Evolving our underwriting capabilities continues to be one of our biggest investments.
We are constantly adding new data sources and analysis techniques so that the picture of the digital enterprise is higher fidelity and more real-time. We follow that with more and more sophisticated machine learning based decision making. Our vision is a future state where we have the ability to underwrite any enterprise in real-time.
To infinity and beyond
This move into insurtech was quite the departure for me. Prior to Coalition, I spent the bulk of my professional career focused on some aspect of monitoring or observability. Whether running Amazon CloudWatch or building network monitoring systems for the likes of AT&T at Tektronix, I have helped to build some of the largest time-series systems on the planet. These were truly big ‘big data’ problems with massively hard problems around correlation, aggregation, and analytics at scale. The challenges here at an insurtech are different but no less interesting and challenging. I am loving the chance to tackle problems that have a different shape than what I have done in the past and make our bit of Coalition history in the world of insurance.
If these insurtech challenges sound interesting and fun to you, we need many (many) more outstanding engineers and product managers to come join our little rocket ship.