October 11, 2016
Quick Takes From Insuretech Connect
A great inaugural event, but to a large extent insurtech innovation is not yet focused enough on the core industry issues.
Last week, I was excited to attend the first Insuretech Connect conference, which brought together entrepreneurs, VCs and industry insiders to focus on the innovative (and some say disruptive) developments within the industry. I wanted to get a closer view of the emerging technology and begin to hear a clearer message about how these developments are connected with the core issues facing the industry, such as: the industry in total very rarely delivers cost of capital returns; the products are complex, and structured in ways that make them not easily consumable by customers; there is aversion to new risks by the carriers given lack of credible loss information used for pricing; a third of P&C premium is absorbed in cost of sales and delivery, an unsustainable figure; etc.
With the event behind us, here are my top takeaways:
1. There are fantastic stories beginning to emerge about the engagement of millennials (notoriously uninterested in insurance products) that over time could be hugely instructive for the broader industry.
Both Trov and Lemonade are genuinely different, with an experience that is more akin to a social media exchange with your friends as opposed to the arduous image (and sometimes reality) of most insurance buying, servicing and claims interactions. Both appear to have genuinely rethought the product being delivered.
In the case of Lemonade, the company has removed the implicit contention between insured and customer with an affinity-oriented dimension: Excess premiums not used to pay claims go to a charity of the customer’s choice. These factors alone (I will cover more below) fundamentally reposition the insurance provider in the mind of the consumer.
Trov is delivering an on-demand, single-item, micro-duration coverage – a genuinely innovative product concept. The takeaway here is that true innovation in customer experience is unlikely if there isn’t innovation in the product. Trov also provides its user with an app that has real value to the consumer independent of the insurance cover — effectively the app is a a super-easy-to-use personal asset register.
The “value in use” delivered in this app is a launch point for an entirely different type of engagement. Metromile is doing the same thing with its free smart driving app, which helps you with where you parked your car, with diagnostics and maintenance and with trip planning. The Metromile app has tremendous value to its users independent of the usage-based insurance the app provides.
So the real question for the industry is whether Lemonade and Trov are just great ingenuity to deliver renters and single-item coverage to a segment that is meaningfully under-penetrated and uninterested in insurance, or whether these fundamental innovations will be harnessed and applied by others not just elsewhere in personal lines but in commercial and specialty lines, as well.
2. Unsurprisingly, the conference was dominated with many who are endeavoring to attack the distribution part of the value chain by changing customer experience and the cost to deliver those experiences. Many of the entrepreneurs are borrowing pages from the countless other categories that have gone through dramatic changes in distribution (financial services, travel, etc.).
It is early days, but I look forward to companies such as Embroker, which is legitimately trying to re-create the entire customer-broker experience (focused on the more complex middle-market commercial risks), with technology as a critical enabler.
One far narrower example is Terrene Labs, which is a really interesting play on big data that potentially flips the application-for-insurance process for commercial insurance on its head. Effectively, the company is developing the technology that combs the public domain to create a near-completed (and far-higher-quality) insurance application based on only a handful of questions. I highlight this venture led by the ex-CIO of Great American as he is seeking to improve the customer experience in small commercial while simultaneously slashing the front-end agency cost of entering the application data to carrier’s on-line systems.
I suspect that the much-anticipated launch of Attune, the initiative backed by Hamilton-Two Sigma-AIG, will feature this sort of change in experience. I anticipate the developments next year on distribution are going to be far more robust and measurable.
3. While there is an intensifying discussion about the Internet of Things (IoT) and the exponentially increasing data that can be accessed to evaluate risk — including sensor technology that can convert risk taking into a continuously monitored, pay-as-you-go model (even in liability classes) — most of this is futurist stuff. The exceptions are usage-based insurance (UBI) in auto, some modest developments in smart home and increasingly smart machinery monitoring you find in a variety of commercial applications.
Yet one company really stood out in its ambitions. The company, Understory, has been installing micro weather stations (wireless, solar-powered, etc.) to get a far more finite view of rain, hail, wind, etc. than the National Weather Service can provide. During a panel discussion, the CEO noted that the company can put 60 of these micro weather stations in a city for the cost of a single large radar system (around $200,000).
It is difficult to cite the specific loss to the industry of straight-line wind and hail (it runs in the tens of billions of dollars in the U.S. alone each year), and hail loss is notoriously difficult given the sometimes long tail to discover it and, in certain cases, the high fraud rate and difficulty to empirically verify whether a hail storm that occurred during a specific period of insurance coverage caused the damage.
But the sort of innovation occurring at Understory was one of the few focused on a core aspect where the risk takers can improve performance and meaningfully reduce loss costs. This is not to say that the many excellent developments around machine learning and predictive analytics applied to underwriting and claims is not similarly attacking these sorts of costs, it is just that Understory is unusual in that it is a tangible quantum improvement in data that can drive improvement in loss costs.
Look out for the next wave of “Understories” and to more tangible results from the variety of vendors pushing the machine learning/big data angle for both claims and underwriting,
4. I finish with my “not so impressed” takeaway. The most obvious aspect missing at the conference was a good economic understanding of the insurance industry by many of the entrepreneurs selling their wares. In some cases, including panelists, they were flatly wrong in their assertion and some showed little regard for the facts.
Even Daniel Schreiber, the CEO of Lemonade (whom I found to be thoroughly entertaining, insightful and articulate about many things, including behavioral economics), responded to a query from the interviewer/moderator in a way that indicates that some independent research suggests that the pricing of Lemonade’s product is a fraction of competitors. Schreiber suggested that the 25% cost for distribution (I interpreted this as total commission) and 40% total operating costs for the industry, compared with the “20% management fee Lemonade charges its customers,” is a key contributor to the difference in costs.
Underlying Schreiber’s comments was an obvious point that the cost of today’s insurance product to the customer is far too high and that innovation has to drive down costs for the insurer and prices for the consumer. At least Schreiber took on the issue in a thoughtful way.
Unfortunately, though, the 25% and 40% numbers are simply wrong. I go back to the factual economics of our industry. The INDUSTRY IN TOTAL DOES NOT EARN COSTS OF CAPITAL, so the industry in total is not getting paid for the risk it is taking. In 2015, 31% of premium (not 40%) went to sales and service. In personal lines, the numbers are far lower. As a reference point, Progressive’s total expense ratio is just under 20%, and Travelers homeowners expense ratio hovers around 28% (with a large part in commissions, given their retail distribution model ).
I am not suggesting that the industry is not ripe for some disruption, but that those are seeking to disrupt (or even enable) it need to understand the macroeconomics and then follow the money (kind of what Understory is doing).
Back to Lemonade. I can imagine that the company has built its infrastructure in such a way that the investors will get an appropriate return from the 20% management fee. I can further imagine that the model may self-select a better class of renters than the wider population and that maybe the fundamental proposition reduces fraud-driven loss costs, so a far lower price could be justified. Yet only a few of those at the conference started with a good foundation of industry and value chain economics, an understanding of the unique regulatory and product attributes that will remain for the foreseeable future, and where and how underwriting and loss performance can be improved.
As these issues come into focus, I suspect that the innovations will begin to fulfill the expectations that are building in the insurtech space.