Itai Ben-Zaken is the Co-Founder and CEO at Honeycomb Insurance, a digital real estate insurance pioneer providing right-priced insurance coverage for condo associations, HOAs, and multi-family landlords and developers. Itai was interviewed by Michael Fiedel, a Managing Director at InsurTech Ohio and Co-Founder at PolicyFly, Inc.
Why do you think commercial insurance lines are such a huge opportunity?
“I believe commercial lines are the next frontier. Innovation in insurance started with smaller policies within much bigger markets. The innovation flow in these markets started from simply making online, real-time quoting available and has moved into leveraging real-time data. Commercial lines traditionally deal with larger policies, which are more complex to underwrite. Given the complexity and the ‘low-hanging fruit first’ approach of most insurance players, commercial lines have been on the back burner for many of the incumbents until today.
So, commercial is a new frontier, and there's a lot to solve there. Whereas in personal lines, at least for the bigger markets, it's definitely a more mature digital market.”
How are you leveraging innovative data-driven technologies to build a unique offering in the real estate insurance space?
“Our team at Honeycomb is leveraging this on multiple levels. First, on a more simple level of digitization, the ease of use has just not been there for this segment of insurance. It's rising, but you still see a lot of small commercial markets do everything on paper. The second part, maybe the more interesting piece, is that we are using big data and data-driven technologies to take underwriting further in terms of sophistication and segmentation, which is pretty tough to do efficiently and cost-effectively without rich data and technological refinement.
You're using very experienced people, especially in cases where it's small commercial. When you're reviewing thousands of risks every month, it becomes tougher. So, we built a much deeper underwriting approach, leveraging data-driven automation and deeper segmentation to ‘right-price’ risk without the need to have an ‘army’ of underwriters. We don't use just the traditional underwriting parameters like age of the building, type of construction and zip code. We go much deeper into providing unique discounts for customers who show us that they maintain the property and have pride of ownership. These unique features of our approach make it a more attractive product for the customer, while making the life of our broker partners easier because they have a more attractive product to sell and in a fraction of the time.
To sum it up, there are two main areas where we're using technology in a unique and innovative way. The first area is streamlining and making everything easier, more accessible and more transparent for the customer and the broker. The second area is making sure that the underwriting engine behind the scenes uses the full power of data-driven technologies that we have in 2021 and deepens the ability to underwrite precisely.”
What has made some of these advanced technologies more practical and deployable in the last five years?
“Google has done an amazing job in making AI development tools accessible to developers. There have been breakthroughs in neural nets in the past 5 years with the ability to use them for analysis of imagery, which many more researchers and developers can use now. But the devil is in the details; you need to know how to use the tools, and in many cases, you still have to develop proprietary algorithms.”
What was the reaction of reinsurers and potential partners when you initially started pitching this strategy for real estate insurance?
“Initially, there was a lot of interest from both reinsurance and partners because it's a market that has traditionally not been very well served. It's clear that there’s a big opportunity here by creating the best product and being the market leader. Because we insure properties that on average are 20 times more expensive than single family, there’s a lot more scrutiny on us to answer the question, ‘How do you guys actually do that?’.
We must be transparent about our depth of underwriting. It's not just about streamlining or creating a digital process where somebody clicks a button and gets a quote online for a $200 policy at massive scale. It's more of a situation where we have the burden to prove that we can do the underwriting in a deeply segmented way and based on a lot of experience in the industry. My partners in this company bring a lot of past institutional knowledge and experience. We fuse that together with the innovative technologies that we bring as an insurtech company.”
How important do you think it is to blend people from outside the industry that have a technology background in combination with those that have more of a deep rooted upbringing in insurance?
“I think it's critical. You want to innovate, but it would be foolish to ignore what experienced property underwriters have learned by trial and error in the past 40+ years in the market. At the same time, we're coming to innovate in an industry that has been pretty stale, so there is a lot of opportunity.
So, if you can take that institutional knowledge and mount it on a platform that's more consistent and less expensive, then you create that unique, synergy of values. You want to be leveraging every little bit of experience that exists, and there are good ways and bad ways to do it. One of the challenges is just figuring out how you get a lot of this great data out of segregated data silos into one place where you can now analyze it and leverage it for real time underwriting. I think that's one of the bigger challenges that some of the incumbents are still facing. Although they want to drive innovation, they operate entirely on legacy systems. One of the advantages that we bring in building on top of a new platform is that it allows us to act on the fusion of data and institutional knowledge much more quickly than others. I've learned so much by sitting, listening and having very passionate debates from both the insurance and technology sides of our business. We make so much progress by bringing this diversity of experience to the same table.”
What misconceptions about technology have you run into while building this business from the ground up?
“One of the things that I've heard or seen is people thinking that they can completely replace underwriters with AI, and I don’t believe this is going to work. I think we're augmenting deep underwriting experience and a lot of lessons learned with technologies that can do the atomic operations more efficiently. The other misconception was that simplifying the process will undermine the level of underwriting accuracy. It's exactly the opposite because the existing price modeling prevalent in the industry is actually quite shallow and not very advanced. A lot of the decision-making is eventually being pushed to a human underwriter who makes an individual judgment on every risk, and that’s both very costly and not always accurate.
We've been taking those pretty simple six or seven data point models everyone uses in the industry and moving them to be 100 or 200 data point models, which gives the underwriting a lot more leverage and relies less on having somebody go in and apply a subjective judgment call on every four-unit building. By creating a system that uses technology and teaching the system everything we can from prior underwriting experience, we gain both increased accuracy and cost effectiveness, and that’s how our customers win.”