Q&A with Adam Kornick

Insurance wasn’t the plan for Adam Kornick, but he ended up interviewing at Progressive. The former president of insurance at Porch Group, who also spent time at Aviva and Allstate, recently sat down with us for an interview.

Share your insurance background.

I studied engineering at MIT and never thought of insurance, but an alumni there encouraged me to interview at Progressive and it was a great place to learn how to innovate in product and distribution while consistently meeting profit targets. The TAM in insurance is large and the regulation makes strategy interesting, which are both exciting. My last role at Progressive was building out what was then called big data and I was able to continue working on how to apply data and AI at Aviva, outside the US, and then at Allstate when I returned to the US.

More recently, I joined Porch as a private company to build and lead the insurance businesses and to take the company public, which was a really valuable experience. Since leaving Porch a few months ago, I’ve been focused mostly on advisory and board work for more mature insurance businesses and working with insurance adjacent startups, including making investments into the cap table. Insurance is my home so I’m looking carefully to find the right fit for my next full time role.

You’ve been experimenting with big data and analytics long before the recent trends. What’s different today?

Software and compute are so much more available (and better) such that delivering AI is much easier than 10 years ago. When I started we used closed source proprietary code to implement algorithms in our own data center which is remarkably different from simply calling an LLM in a browser or via python with a few sentences of text. In the same way that the internet changed the cost and richness of communication and mainframe computers 30 years earlier changed the cost of calculation, AI will lower the cost of using unstructured data. It will take time but tasks that are human centric today, like reading a policy, sorting and interpreting piles of business specific documents or creating marketing creative, will be automated and become more and more digital and personalized over the next 2 decades.

We can’t ignore the significant generative AI developments. What are the main Gen AI use cases in your view for insurance companies that have limited interactions with their customers?

Any unstructured data is a great candidate to get better outcomes at a lower cost with generative AI. Even with limited customer interactions, an insurance company has unstructured data. If an insurer has complex risks where human underwriters can take advantage of offloading tasks to gen AI that will give them an advantage over time. Same thing applies in a different way to healthcare payers and the complexity of deciding how much to pay and what to cover. Even companies that don’t have many consumer interactions have plenty of internal, agent or other partner interaction and exposing information through a gen AI interface, like a chatbot, can give agents, for example, answers faster and at a lower cost. Finally, I think about all filings that have to be generated and analyzed with regulators (typically by state or province in the US and Canada). Getting those filed correctly and quickly will help everyone.

You’ve been with major insurers and in the past few years you led a new insurance venture at Porch. Many criticize new insurance ventures about falling short when it comes to insurance fundamentals. What are the main elements that are making this space challenging to newcomers?

Many successful entrepreneurs have built multiple businesses, most often in software or consumer products but the intuitions from those businesses have to be re-examined for insurance. Most insurance businesses are some combination of highly regulated, variable cost, and/or capital intensive. That means that intuitions that often work well in other industries like growing rapidly and unprofitably with the plan to capture margin in the future require different planning. It’s also been a tough few years, with a pandemic, wars, inflation, and high levels of natural catastrophes. Newcomers would have had an easier time if we picked a five year span when those things didn’t happen. Conversely those who have made it through may be on to something. Remember that Progressive and Allstate where once startups, and Allstate was launched inside Sears and had to grow its insurance acumen.

What does the future hold for insurance in the US?

Sales channels will continue to evolve as embedded insurance grows but the agent channel will continue to matter even as other channels grow. In Life, the asset manager driven roll-up will probably continue and I suspect regulators will spend a lot of time trying to design the right framework for this shift. In Property and Casualty we’ll continue to see the technology component of cars, like sensors, lower frequency but increase severity, and I think telematics will continue to grow the way financial scoring did in the early 2000’s. I think we’ll also see higher property replacement costs and more severe weather continue and this will lead to higher prices and higher deductibles for consumers and capital needs for insurers and reinsurers. This has already driven innovation in the E&S property market, in the use of structures like captives and exchanges, and in the growth of collateralized reinsurance. The use of more and varied data will also continue, that’s been a big part of my work at every company. I also think we’ll see big differences by state: states that don’t permit insurers or reinsurers to earn the right return on capital will see that capital move elsewhere and a loss of availability.

In the first few years of insurtech, full-stack carriers attracted most of the attention and high valuations. Now it seems that brokers have become the main event. Has distribution become more important than product innovation?

Distribution has benefited from technology-enabled sales processes and from the higher commissions that have come as higher loss costs have driven premiums higher. The technology enablement will continue, but the loss costs will filter through the system and impact everyone. Businesses with capital are not going to let others spend it indefinitely, especially when the risk free rate is so high and balance sheet companies are competing with higher expected returns from investors. The balance sheet companies are going to innovate in product and distribution so that they provide advantaged access to the distributors that make them (the full-stack carriers) profitable and that’s going to swing the pendulum the other way. There will be exceptions but that will be the cycle.

We’ve seen some major brands such as Amazon and Intuit get into insurance only to exit after a few short years, and many insurtechs have been struggling for quite some time. Will we see insurance companies that can actually challenge the major players in the space?

Sure. Companies like Progressive (took off in the 80’s), GEICO (ex-USAA employees started it before Berkshire Hathaway started to invest and grow it in the 70’s), and Allstate (only spun out of Sears in the 90’s) all came to dominance in property and casualty in the US relatively recently. Those businesses have real moats from regulation, the cost of capital, and the cost of a brand. They say history doesn’t repeat but it rhymes so the geography, business model, and lines of business will vary but it will happen over the time scale of decades. It can take many new entrants but eventually one takes advantage of speed, a new business model, product innovation or something else to gain dominance.

Which insurance/insurtech company has been performing well in your view?

Among carriers, Progressive’s skill in pricing auto and distribution sophistication stands out to me as does Chubb’s ability to deploy capital to the best return/risk profile between their many geographies and business. Both have a strong focus on writing every unit of business at a profit (or at least trying). In software, I think Snapsheet continues to be successful at innovating the claims process and providing a configurable cloud platform that brings claims innovation to carriers more quickly and easily. I think Goosehead has been disciplined about using technology, a franchise business model, and a focus on customer retention to get better outcomes than many other distributors. All of them seem to be focused on measuring what they are good at and doing more of it and insisting that things that don’t work get improved or shut down.

Which non-insurance player do you believe should enter the space?

Companies with access to customers (businesses or consumers) and the data to price accurately with low effort are the candidates. A couple that fit the bill are rideshare companies and banks. Rideshare companies have access to millions of their own drivers and the mobile phone app data for pricing auto insurance. Banks have millions more consumers and businesses and the data to setup the right coverage at an accurate price with little customer effort across many lines. I think they haven’t because it requires different capabilities (building a nationwide claims infrastructure is challenging) and a new regulatory structure. They are focused on their core markets and if it was easy it would have happened but each of them should be looking for ways to explore insurance both to accelerate their core business and to build a new one.

What advice would you give to founders building new insurance ventures?

Well the obvious stuff always applies, be humble, hire great people and have a clear proposition to make things better for customers but above all else, make money. Revenue growth that exceeds the cost of capital unlocks so much in a regulated financial services business. Unpack your intuitions and figure out how to meet your investor return metrics without growing rapidly at a loss planning to make it up later. The closer you are to the risk the more true this is. I would also say make a reasonable chunk of the great people you hire those who know the industry. I’ve seen in my own experience that entity structures like captive, reciprocal exchange, and stock company and business model choices on whether to deploy product innovation as a carrier, MGA, agency, or software provider deliver meaningful value and open up access to capital. Some companies burn too much money and time acquiring knowledge when they could have had employees or board members who had learned it in prior roles.

What would you do if you didn’t have a career in insurance?

Engineering or running a software business. Before insurance, I worked as an engineer at a coastal engineering firm and at Ford and loved the challenge of figuring out a car engine or an oil drilling operation. I also worked directly with technology early in my career and learned a lot about running great businesses beyond just having a great product.