Insurance Startups: Digitally Native is not Enough
Insurance is the hottest sub-sector of fintech these days. This is driven by a number of very real trends that make meaningful disruption of the insurance industry possible today: terrible industry NPS, lack of engagement from the next generation of buyers, lack of product innovation, and reams of new data available for underwriting, pricing and targeting risk. Given this confluence of trends, one of the most exciting investment areas is the opportunity to re-imagine existing insurance product to build a new insurance carrier that resonates with the next generation of insurance buyers. This means either insuring new emergent risks or fundamentally changing existing product. As we think about opportunities to build new carriers in insurance, there is a lot that we can learn from the consumer cost savings and customer experience improvements that online lenders pioneered in the past decade.
It is easy to assume the value of LendingClub, Prosper, et al is driven by the P2P model, but the truth is more nuanced. P2P lenders, or really marketplace lenders (MPL) as it evolved to, have grown rapidly not because of P2P, but rather the economic surplus that the model creates for borrowers. LendingClub and Prosper save the average consumer ~700 BPS on a loan they refinance through those platforms (say taking credit card debt costing 20% and reducing it to 13%). They can do this by reducing significant cost in the system and superior risking. This is a true economic advantage. The fact that it can be transacted purely online is the icing on the cake. The lower cost drives a lot of demand, and institutional investors follow to where there is high quality risk-adjusted yield. MPL is not what drives the loan volumes, but rather the loan value is driven by the economic surplus for the borrower that MPL creates.
When we look at insurance, the same opportunity to create economic surplus exists by leveraging the myriad new data sources available. New data can provide huge value in not just the underwriting process but also in the quoting and prospecting processes. In the life insurance market, leveraging new datasets via API could allow for a much more streamlined and accurate risking and pricing process. In homeowners, CoreLogic offers a huge amount of data to underwriters, but it is often inaccurate requiring redundant on-site visits; if you could acquire more accurate data about a property at the time of quote, insurers could price risk at scale with far more accuracy and underwrite quicker. With aerial photography, drones, IOT, and machine learning this is doable, and leaders in the industry are clamoring for these abilities. All of these data sources will make it easier for the industry (disruptors or incumbents) to offer a better price to the best risk classes. The challenge for incumbents is that their IT systems are not set up to consume these data via API.
The previous point notwithstanding, MPL also offers a significantly improved customer experience. While the press may like to make a big deal of MPLs online-only distribution, the truth is that underwriters at incumbent banks are completely removed from the borrower as well — they never meet the person face-to-face. For instance, a Bank of America card customer with direct deposit into their checking account must still resubmit all of their personal information for underwriting a limit increase — despite the fact that the bank already has access to all of this data. Getting a new unsecured loan, or refinancing student debt is even more painful. MPL reviews the same data, but consumes it more efficiently, and processes it faster such that fully underwritten decisions are returned to the customer in days rather than weeks or months. In venture capital, we often say that if you’re differentiating on UX you need to have a 10x improvement in experience — this is hard to measure, though when you experience it often feels like “magic”.
This brings us to insurance innovation. There are many people attempting to build a “full-stack” insurance company. These new platforms are often digitally native and promise a significantly improved customer experience. The challenge here is that outside of health insurance, few insurance verticals have multiple touch points a year with the insured where these improved experiences can create value. Outside of claims events, life and P&C (home, auto, etc.) are mostly set it and forget it policies. Thus in these markets the biggest opportunity to improve experience is around the initial acquisition and underwriting of a customer. Insurance is brutally competitive from an acquisition point of view and conversion funnels often see greater than 50% drop-off from the initial touch point with a customer to binding a policy. While the current agent-based distribution model may seem antiquated (it is) and sub-optimal to attract the next generation of buyer (it is); there is a good reason why agents get ~80% of first year premium — its hard work to source and bind high quality risk. The big opportunity for insure tech is not to make the existing insurance products digitally native, but rather to fundamentally alter the product or its underwriting process to create an order of magnitude improvement to the customer’s experience.
Ultimately, to meaningfully acquire customers, innovative “full-stack” insurance companies, like online lenders, need to offer a better price and experience as compared to incumbents. If a new entrant is only price-competitive, they will need to offer a 10x improvement in customer experience. The good news is that there is a lot opportunity for innovation here and the market is massive. Insurance companies have left a yawning gap when it comes to customer experience, and a multitude of new data sources will undoubtedly complement the current underwriting process. The challenge for insurance startups is not to copy lending, but rather find the “magic” in this market that fundamentally changes how insurance products are purchased and sold.