Technology and insurance

Several innovations are getting plenty of hype to disrupt the insurance industry. We've seen this before: designs like telematics were considered groundbreaking fifteen years ago. We have yet to see it thoroughly disrupt the car insurance market, as many of us imagined in the early 2000s. Would this time be different? Or is all insurance hype the equivalent of dot-com boom and bust in the 1990s?

Previously, I highlighted the innovations most likely to disrupt the P&C insurance industry. A mixture of any or all of these emerging technologies, combined rightly, could accelerate the industry's speed of change, even to the point of fundamental disruption and a new paradigm. These technologies could reinvent the risk transfer process. I claim insurance is an ideal digital commodity, and I'm not alone in that view. Why is this so?

• Insurance requires no physical assets, including production plants.
• Insurance requires no complex multinational supply chains to transfer physical goods.
• Insurance doesn't even need a vendor!
• Use a few million dollars to launch an insurance firm.
• The market is so big, and even modest success will contribute to sustainability.

More than a digital world

Direct writers in the US and a highly competitive and creative market in the UK and Australia point to what's possible. For example, innovative microinsurance products bought and serviced on smartphones in India illustrate this. Industry trade publications like Digital Insurance, Coverage, and others point to traditional players and startups' achievements in exploiting insurance to streamline procedures, eliminate consumer pain points, sell more plans, and create new products and services. Bottom line: insurance should do much more in the digital realm.

If more can be done in the digital world, why hasn't it happened yet? Entrepreneurs and visionaries can express various potential risk transfer paradigms, including blockchain-based, decentralized peer-to-peer (P2P) insurance. An API platform with modules for claims servicing when required and loss prevention may promote this mechanism. Another idea is that car manufacturers, and probably even structural engineers and construction contractors have so much experience with their structures and goods that they are more likely to take on insurance liability than conventional property-casualty insurance.

P&C insurance can be similar to life insurance, where you pay a certain amount per month, but some are not used for claims accruing to you. Your deductible increases under this model, creating a nest egg over time that will help you handle the financial effects of failure without leaving years of "unused" premiums. The explanation that these systems and many more do not exist today—or, if they do exist, have not gained widespread acceptance and counterproductive to the existing status quo—is due to entry barriers.

Any new risk transfer paradigm must overcome three barriers:

1. Confidence in the risk transfer (insurance) ecosystem
2. Using historical data and retrospective analysis to assess premiums, keep loss reserves and fixed capital thresholds
3. Missing customer interaction

Current model

The current model is focused on 300 years of building insurance product trust. They collected upfront premiums for potential contingent losses based purely on a pledge that generates all kinds of perverse incentives. Indeed in the early days of insurance, there are several tales of swindlers, robbers, and crooks running Ponzi schemes to defraud people from their capital. But well-intentioned insurance companies can quickly come up short if they don't charge actuarially sound premiums and keep enough money in the reserve to cover all future losses.

Many institutions, laws, and procedures required to create trust in the system:

• Developing actuarial science to know how much to charge in reserves.
• Establishing insured funds sponsored by government agencies when an insurer goes bankrupt
• Creating regulatory authority to regulate carriers and train and license agents to do business.
• Contract compliance is legally binding contracts by the court system.
• Developing compensation theory.

The existing system has developed to amend claims (by qualified and certified staff) that investigate allegations and ensure that claimants are paid what they are owed—no more and no less—to restore them to their pre-loss condition (less any deductibles that might apply and subject to policy limitations and exclusions). All these buildings were built for one primary purpose: building trust between two parties. These parties are frequently unfamiliar and involved in a large financial transaction on both ends. Confidence is encouraged by daily premium payment by insured and covered claim amounts of payment by the insurer. Any new risk transfer model to replace the current order must create a similar level of trust.

New ventures

Without it, there's no chance of disrupting the industry in any meaningful way. Uber and Amazon find ways to disrupt conventional businesses like taxis and department stores. They used conventions like user feedback, monitoring rides or shipments, and direct customer contact. Via applications, messages, email, and phone have developed enough confidence that people are willing to replace them (relatively new startups and gain advantages over traditional players. Millennials tend to be more trusting than traditional, stodgy businesses.

Startups in the insurance space need to create the highest degree of trust for their customers. A second barrier relies heavily on historical data to market good insurance policies. Established players have a considerable advantage over startups because they sit on massive data troves. In a Big Data environment, predictive analytics, artificial intelligence (AI), machine learning (ML)—a world run by algorithms—data is a simple commodity required as input to run a successful company.

Regulators typically need historical evidence to validate rate adjustments and guidance. Rating agencies and reinsurers also like historical data to cover their rating or reinsurance. For entrepreneurs, collecting data is a significant obstacle. Companies that produce sensors like telematics devices and smart home IoT sensors that stream a continuous data feed that can directly observe activities need claims and loss cost information to build their value propositions. Offering new insurance products and services is difficult for businesses without data. While you can launch when you have the capital to back you, heavy initial losses would displease investors. It's a challenge not to underestimate how to pool risks into the correct segmentation scheme. Having all this historical data, though, isn't always an advantage. Store and wrangle the data in a format that can be useful for analysis is a huge undertaking.

Many legacy transactional frameworks are incredibly complex. A decade or more has been spent building massive data marts and data warehouses to store, view, and report dashboards effectively. Data lakes are the new favored AI feeding process. All those old flat files that corporations spent millions converting into complex data structures are almost back in vogue. Existing players also don't collect data that can be easily accessed and used. This is particularly true for unstructured phone calls, photographs, and text descriptions.

Customer interaction.

The final obstacle to conquer is customer experience level. For P&C insurance, the customer mentality is usually set and forget it." Insurance is required to accomplish some greater purpose, such as driving a vehicle, renting a property, or buying a home. Insureds and insurers have no day-to-day contact.

Insurers.

Insureds are unlikely to notify their insurer unless they claim to report exposure changes (e.g., a new car or driver, moving to a new home, etc. or premiums escalate dramatically. On the other hand, small business insurance is more involved and needs an assistant. But ongoing contact is impossible. For larger accounts, insurance is part of risk control. Risk management practitioners are working with a broker on a package that includes insurance coverage, liability assessment, and loss reduction techniques. This minimal interaction model is the pattern that customers and insurers came to expect for good or bad. Are these the engagement levels insured want?

Carriers and entrepreneurs aiming to deliver high-touch services, such as apps scoring driving trips to the store or reports, house water flow rates, and a lot of data to help people navigate their everyday lives, the question is how customers will respond? Would they enjoy higher engagement? Or have insurance companies not won the same conversational intimacy as Apple, Google, and Amazon products? And do customers value this knowledge beyond its novelty?

A teen driver's parent can appreciate input about how well or poorly their trip to the movie theater went, but what about their morning work commute? Can they use the "feedback" to alter or continue driving habits the same way? I've seen tragic losses over and over again in my twenty-year industry career - the injuries caused, life changed forever. And even though those affected gave testimonials about their harrowing encounters, most people had a cognitive disconnect. Those not personally affected by tragedy feel sympathy for the person who suffered a significant loss while remaining resolute that It will not happen to me."

Human psychology is tuned to be unrealistically optimistic about the risk of failure. People assess the effect of low incidence, yet high severity events can have on them. And how's your value proposition pitched? Are surveys and interviews a credible indicator of what customers want? What issues are they trying to solve? What are their pressure points? It can't be about seeking use cases that validate the technology solution. Companies must recognize the real challenges faced every day by real people living whole lives. Most significantly, people must view these as actual issues before they value solutions.

Tearing down barriers.

How to conquer these barriers? Several promising technologies allow utterly new risk transfer paradigms. Many have an insurance feature. Others could be more or so radically different quasi-insurance items we won't recognize as insurance.

Trust-building technology:

• Chain block
• Peer-to-peer (P2P) technology, like social media, etc. Data access-transforming technology:
• Telecommunications
• Internet of Things (IoT) like home technology
• Language Processing (NLP)
• Intelligence (AI) like machine learning (ML)
• Aerial photographs