Tag Archives: evan greenberg

Data Prefill: Now You See It, Now You Don’t

At children’s birthday parties, a guest magician may utter the well-worn phrase, “now you see it, now you don’t” – and a bouquet of flowers disappears. That trick, a heartwarming memory for many, also relates to the vast quantity of questions on an application for commercial lines insurance.

It’s daunting for a business owner to come face to face with the numerous blanks on an insurance application. Much of the required information is not immediately at hand – or not understood at all. For distributors, the familiarity with the content is certainly there – at least for seasoned personnel. But the time it takes to fill empty boxes keeps them away from more useful interactions with customers. On the other side of the transaction, company underwriters need information to price the risk. For a very long time, the industry has been at a stalemate.

A conundrum? Not any longer.

See also: 3 Keys to Selecting the Right Platform  

Enter data prefill and new data sources. Data prefill certainly isn’t new – personal lines insurers have employed it for some time. But, the impetus to use the capabilities in commercial lines has not been present until now. Business owners require a simpler application process, and distributors need to be freed from clerical tasks. Undertaking a data prefill initiative may be a simple decision for some organizations – but for others it may be a challenge. In either instance, SMA has a five-step analysis process (Why, Who, How, Where and What) that can guide any organization looking at data prefill. It’s important to approach the initiative with a measured assessment to ensure a successful outcome, even if everyone is already on board with data prefill.

Given the press that organizations such as Cake Insure and Pie Insurance have received, it might be easy to assume that data prefill is all about small business and workers’ compensation. Clearly, there are significant opportunities in the small business arena to condense insurance applications down to three, four or five pieces of data. Evan Greenberg, CEO of Chubb, has declared that the current 30 questions in small business applications will be condensed to around seven within 18 months. However, it would be a mistake to assume that data prefill is just about one commercial lines segment.

In fact, insurers covering all but the most complex jumbo commercial lines have an amazing opportunity to use the same data integration techniques for data prefill to automatically integrate data into more complex lines of business – to improve data accuracy and thus drive profitability. Regardless of the line of business or size of the business insured, augmenting application data with new, emerging data can support underwriters in their decision making. And, perhaps, it can eliminate the need to obtain information from business owners and distributors and promote a much greater degree of accuracy. SMA’s recently released report, Transformation in Commercial Lines: The Five Steps for Data Prefill, provides a view of this.

See also: The Problems With Blockchain, Big Data 

This brings us back to “now you see it, now you don’t” and the disappearing questions on commercial lines applications. Having spent a long time as an underwriter, I recognize that it is unsettling to think about losing the data elements that one has relied on to make decisions. However, with data prefill, that data can be found and used in many ways: eliminating questions on applications for small businesses and prefilling internal systems for more accurate decisions on complex lines. No one will be deprived of data – the source will just be different – an insurance magician’s answer to several challenges!

An Insurtech Reality Check

If you’ve got your eyes set on technology that won’t move the needle this year, it’s time to reevaluate what can provide bottom-line results in the short term. AI and machine learning will have their day in commercial insurance. But what are you doing today to drive tangible business results? Insurtech does not have to be a “pie in the sky” endeavor. It can be deployed right now.

Just a year ago, the insurtech conversation was all about innovation labs, blockchain, IOT, wearables and, of course, AI. Now, the dust has settled a bit, and the realization has set in that those bright, shiny objects may take years to make a real impact on re/insurers’ bottom lines. While they are still undoubtedly vital to innovation, long-term success and survival, it’s important to strike a balance between “pie in the sky” and practical. Last year’s devastating catastrophes served as a catalyst for more focus on short-term solutions that can improve bottom lines—now. Not years from now. This swing to here-and-now solutions was recently articulated in an article by Ilya Bodner, founder of insurtech startup Bold Penguin, where he notes:

“Insurtech is moving rapidly now into commercial lines where the attention and intent is focused on solutions that will deliver a strategic and immediate return on investment (ROI)….Insurers are moving away from bright, shiny, insurtech objects and toward service partners, emerging technologies and solution providers with a return on investment more immediate than promised for five years down the road.”

I second this sentiment. P&C risks are changing, as evidenced by 2017’s $144 billion in global insured losses and a commercial lines combined ratio of 104%. And, while a strong market made many insurers whole last year, that is not a guarantee going forward. The next hurricane, flood or wildfire won’t wait for you to innovate. Insurers must find ways to bring innovation to their bottom lines now. Don’t get me wrong, pie in the sky is good—and it is necessary. But insurers must strike a balance between their long games and short gains. You need both.

Caution: The hard truth

I don’t have to tell you that following last year’s back-to-back hurricanes there was an outcry about how the models got it wrong (of course, it didn’t help that some modelers put out early and grossly inaccurate estimates that incited market confusion and concern). Here’s the hard truth: Insurers also got it wrong. Got it wrong by using a single view of risk; by not taking advantage of innovations in data; by taking too long to operationalize data; by waiting for the perfect, utopian platform (in-house or commercial) to be built or delivered; by expecting legacy analytics software to deliver the scalability, reliability and insight required to act efficiently and effectively. No longer can insurers approach risk The. Same. Old. Way. Risk is changing. You must change with it. And the good news is, integrating insurtech in a way that helps you better assess and manage the evolving landscape of catastrophe risk doesn’t have to be time-consuming or costly, and it can produce immediate results.

Here are a few of the challenges that insurers face that insurtech can help them address, in the here and now:

  • Reality: Models provide a “framework for thinking; they don’t represent truth.” Evan Greenberg, chairman and CEO of Chubb, recently stated, “Given there have been three one-in-100-year floods in 18 months, how can Harvey represent a 1% chance of occurring, as the models suggested? Models provide an organized framework for thinking; they don’t represent truth.” Now, we all know models serve an important purpose, and our clients can derive insights from modeled data within our platform. But models must be taken with a dose of good old-fashioned human judgment. Models and the outputs are nuanced. It’s all about identifying the right models and model components that best represent your lines of business, geography and business practices. But it’s also about balancing resources and business value with this expensive exercise. You need to have an intelligent conversation about model nuances—and figure out the “so what” questions that models provoke but don’t answer.

See also: Can Insurtech Rescue Insurance?  

  • Reality: You can’t handle all the data. There’s a gap between the wealth of data now available and an insurer’s ability to quickly process, contextualize and derive insight from that data. Insurers are generally frustrated by a lack of process and an easy way to consume the frequent and sophisticated data that expert providers put out during events like Harvey, Irma, Maria, the Mexico City earthquake and the California wildfires. Beyond the sheer volume of data, insurance professionals are expected to make sense of it by using complex GIS tools. In reality, you have all this data but no actionable information because you can’t effectively make sense of it. Even insurers with dedicated data teams and in-house GIS specialists struggle to keep up. (SpatialKey tackles this problem by enabling expert data from disparate sources (e.g. NOAA, Impact Forecasting, JBA, KatRisk) and putting it into usable formats that insurers can instantly derive insight from and deploy throughout their organizations. We do the processing work, so our clients can focus on the analysis work.)
  • Reality: Your best data is your own, but you’re not benefiting from it. It’s one thing to be in possession of data, and quite another to be able to realize its full value. Data alone has little value. One of our clients, for example, needed a way to re-deploy its own data to its underwriters, so we helped the company integrate an underwriting solution that would put its data, along with expert third-party data, in the hands of its underwriters—all from a single access point that would consolidate disparate sources and drive enterprise consistency.
  • Reality: Your customers expect on-demand; you should, too. Your customers don’t want to wait for a quote or go through a lengthy process to submit a claim. Our society is instant everything, and while commercial insurance may not be held to the same real-time pressure as personal lines, it is moving in that direction. When you need the latest hurricane footprint, you need it now, not four hours from now. When an earthquake strikes Mexico City, you need to understand your potential business interruption costs today. When a volcano is erupting and no drones are allowed in the surrounding airspace, you need a geospatial analytics solution that can help you provide advanced outreach to insureds and do the financial calculations to understand actual exposure. Likewise, when your underwriters are trying to win business, you’d rather they spent their time evaluating the risk than searching for information.

Who knows what this hurricane or wildfire season will hold. The question is, are you prepared to handle it better than last year? What changes have you made to strengthen your resilience and that of your insureds? What has been learned and applied for meaningful results? It’s a misnomer that insurtech and disruption go hand in hand. Some insurtech solutions are built to complement—to drive efficiencies, cost savings and underwriting profitability—not necessarily replace existing processes or legacy systems. Data and analytics is an area where insurers, brokers and MGAs can still improve their bottom lines yet in 2018.

See also: To Be or Not to Be Insurtech  

Take down the pie and dig in

My intention is not to dilute the importance of up-and-coming insurtech technologies, like AI and machine learning. They will undoubtedly help insurers compete as risks become more complex. My point is that those longer-term technological investments must be tempered with an understanding of what technologies will help move the needle in the present. You can strike a balance between pie-in-the-sky insurtech and insurtech that works for you now.

Why #Insurtech Doesn’t Matter

Last week, I included my summary of what we do in Insurance:

Insurance is a business where we provide people with peace of mind, allowing them to know that there will be a monetary solution provided when they suffer a major loss/accident (or minor, depending on coverage purchased). This loss/accident can either in the form of health, death or to some sort of property, and the solution is at a time when a person typically needs it most. That is the core of our business. 

This summary also relates to the three pillars of Insurance, which I mentioned a few weeks ago:

  1. Pricing –  Was the policy I purchased priced properly to take care of the costs of the insurance company running its business, and will it have enough?
  2. Reserves – to pay my
  3. Claims – in a timely manner.

As with many of us, I read and follow a lot of news on insurance and insurtech. Every day, my LinkedIn feed and email inbox is flooded with insurtech news, including new investments in startups, new insurtech partnerships formed, expansion of startups into new markets/states, etc.

I love reading all of this – as it shows the growing level of awareness of how new technology solutions can enhance the customer experience and also help companies with operational efficiency.  I am a huge fan of what the future entails.

However, I am also cautious of the risks currently present in the world of Insurance (and the world in general!).

See also: Insurtech Innovator – CyberWrite  

Currently, the pace of change and adoption of insurtech solutions is faster than ever before. It seems there are no signs of slowing down. However, as with any good plan, it is important to have risk mitigation and contingency plans.

The new technology solutions that we are building for the insurance industry (i.e., insurtech), are just an enabler. It’s not that these solutions don’t matter…. But, if the risks are not managed properly and plans are not in place for these solutions, then the progress of the many insurtech initiatives may slow down, or in some cases, not be around to matter.

What are some risks as it relates to insurtech? I will focus on three, which have been themes in the news for the past couple of months. In fact, these are risks that exist in our industry regardless of insurtech.

By no means are these the only risks that need to be mitigated, yet I do see these as some of the big ones:

  1. Macroeconomics
  2. Weather/Natural Disasters
  3. Regulation

Macroeconomics

Since 2008, global stock markets have been on a tear. It’s no wonder that there is so much money pouring into insurtech investments.

What happens if there is a market correction and we go into another global recession? Will we see the same sort of investment in insurtech solutions as we have been seeing?

And that’s just the equities market. What about fixed income?

In this FT article from August, Chubb’s CEO Evan Greenberg warns about the low interest rate environment and its effect on insurers. He says, “Many companies are not earning their cost of capital — and many are losing money, or will lose money in the future.” This is a big deal. This may have an impact on an insurer’s ability to pay claims in the future. Obviously, insurers will have to keep their solvency requirements due to regulation, but if this continues, we could see massive premium increases for customers and withdrawal from certain product lines.

Stock markets and fixed income aside, the next big risk that could affect the progress in insurance and insurtech has to do with climate change.

Weather/Natural Disasters

Over the past few months, we have seen Hurricanes Irma, Maria and Harvey ravage much of the Southeastern U.S. and islands nearby. California has been blazing in fire. In other parts of the world, there have been many natural disasters, too. I’ve seen a number of articles on this subject. They range from “how to claim from your insurance company in wake of natural disaster” to “how much insurers will be out of pocket for weather-related claims.” With climate change increasing, the unknowns also grow. I’ll admit, I’m not an expert in catastrophe pricing, but I would suspect that this increasing factor will make it much more difficult to price products.

So, equities may fall. Interest rates may not come up. And natural disasters could be on the rise. These risks are big, but the last one could take the cake: regulation.

Regulation

President Trump has signed two executive orders – one that will allow customers to purchase cross state border and one that limits funding for Obamacare (though that has seemed to change course).

The impact that these have on the U.S. healthcare and insurance market is unknown for now. This is a topic that deserves its own write-up, and I plan to cover this sometime in the near future.

Regulation can really screw things up; if not looked at properly. I wrote about government collaboration a few weeks ago. Some governments are more open to collaborating with incumbents to better understand fintech and insurtech. However, for those of us who have worked with regulators, we know that their minds can change quickly, and knee-jerk reactions can be made, forcing our plans to change.

Different product lines have different opportunities and different risks

For some lines of insurance, mainly P&C, insurtech has a huge play, and there are many opportunities to disrupt and change the current Insurance value chain. If autonomous cars come into existence, the whole auto Insurance industry will change. For property insurance, smart homes and devices to monitor buildings will help to better optimize pricing and policies for consumers.

For travel Insurance, insurance to protect material objects (mobile phones, electronics, etc), UBI and insurance for the sharing economy, there will be opportunities to disrupt and enhance the customer proposition, too.

For life, health and catastrophe, it becomes a different story.  We see a lot of term life online, but what about whole life, universal life, annuities, etc? What about other, more complex products for individuals/businesses (disability, long-term care, commercial)?

See also: Innovation — or Just Innovative Thinking?  

My biggest worry comes from within these types of products. My years in insurance have primarily been on the life, health and annuities side. The pricing structures of these types of products have a longer tail than P&C. Health is annually renewable, but the cost of healthcare and frequency of visits to doctors have been increasing, which will make pricing more difficult.

So what can we do about this?

First and foremost, every startup and incumbent needs to have a risk mitigation strategy and contingency plan as it relates to their insurtech initiatives. It is easy to get caught up in the excitement of what we are doing, and talking about risk is not always the most fun. The risks above are just a few macro ones. Each company and each initiative will carry its own set of risks, which need to be assessed accordingly.

Second, collaboration continues to be key. Especially cross-border collaboration. We need to share best practices globally. Regulators will also need to continue to work with incumbents and startups to understand the solutions being put in place and risks to customers.

Third, actuaries need to get with it – quick. They need to use their skills of actuarial modeling and work with the data scientists out there to better understand all the data points available to them and how this can be incorporated into pricing models.

The marrying of actuarial pricing principles and data science will be one of the most powerful forces of change in our industry. Incumbents have been managing risk for hundreds of years. The nature of managing risk has changed with the explosion of data. It’s no longer about just looking at what has happened in the past and predicting what will happen. Let’s also get underwriters in this conversation.

We need to find opportunities to know what is working where, and what is also not working, so we can plan accordingly. We are all in this together, and we need to help enhance our industry together. We all have a collective responsibility, ultimately, for our customers.

This is a repost of my article on Daily Fintech. I look forward to reading your comments on this article and engaging in some discussion.