Totally Changed Industry Landscape Only Three Years Away

This article is part of Carrier Management’s series on the Future of Insurance.

Taking our cue from Standard & Poor’s Director Tracey Dolin-Benguigui, who predicted during an S&P insurance conference that “the property/casualty insurance industry will look very different in the next 10 years than it has in the last 30,” we asked the leaders of new firms and old to share their visions of the industry between 2017 and 2027.

Pranav Pasricha, CEO, Intellect SEEC said it will only take three years for 30 years’ worth of change.

Pranav Pasricha, CEO, Intellect SEEC

Pranav Pasricha is the CEO of Intellect SEEC (www.intellectseec.com), a provider of insurance software with an extensive portfolio covering distribution, underwriting and claims.

He has transformed from being a traditional executive to a disrupter pioneering the deployment of big data, AI and IoT in the insurance industry.

Q: What major changes do you see on the horizon for the property/casualty insurance industry in the next 10 years? What will insurance companies, insurance leaders, the industry and its workforce look like in the next decade? What risks will they insure?

Pasricha (Intellect SEEC): The insurance industry will change more in three years than in the last 30, and the disruption that is about to hit us won’t take that long to change the landscape. This is the “Exponential Age.” Internet business models took 10 years to take hold in insurance; mobility took about three to five years; and AI, machine learning and big data will be even faster.

The combination of big data and AI will fundamentally change the paradigms of insurance in all respects: distribution, underwriting and rating, and claims assessment. Historically, people have filled out forms (earlier paper, and now electronic) to gather risk information, and it was usually done through agents and passed on to underwriters, who assessed risk based on traditional underwriting rules and priced based on statistical models of historic data.

Today, there are thousands of structured and unstructured sources that can provide a lot more information—not just about the risk but also its operating environment, people and business practices that may influence its loss propensity. These may include social media, traditional media, various government and private databases, information from IoT and telematics devices, and satellite and drone feeds. This vast information set can be assessed automatically and objectively by AI-based programs and will produce a much better risk assessment.

“Today there are thousands of structured and unstructured sources that can provide a lot more information—not just about the risk but also its operating environment, people and business practices that may influence its loss propensity…This vast information set can be assessed automatically and objectively by AI-based programs and will produce a much better risk assessment.”

Further, using machine learning algorithms, we can predict loss propensity across thousands of factors which can rate far more superior than the statistical models we currently use. Similarly, loss detection and assessment can be more immediate and accurate using IoT data and machine learning models, fraud detection can be more precise, and payments can be immediate.

Insurance companies are still dealing with a lot of processes that require human intervention and judgment. Over the years, we have taken paper-based processes and digitized them in many places, and some automation of basic tasks have evolved. However, it is still one of the most labor-intensive sectors of the services industry. In the future, we envision significant elimination of these basic processes and therefore a reduction in the headcount of insurance companies—perhaps down to about one-half to one-third of the current workforce. For example, AI and big data-based front-end platforms can eliminate or significantly reduce the need for agents and underwriters for all but some of the most complex products. Robo-advice platforms can do much better holistic risk and needs assessments to recommend products. Big data platforms can virtually eliminate application forms and the whole process of checking and assessing them. Machine learning-based tools can do automated underwriting.

“Leaders will need to become significantly more technology- and data science-savvy and will need to manage a smaller but more specialized set of individuals who will expect a very different form of organizational alignment and motivation.“

In this environment, most of the skillsets required will be data scientists (as against actuaries), UX and visualization and machine learning experts (as against agents), and some deep domain and risk engineering specialists (rather than traditional underwriters). Leaders will need to become significantly more technology- and data science-savvy and will need to manage a smaller but more specialized set of individuals who will expect a very different form of organizational alignment and motivation.

The products offered by a large part of the traditional insurance market are set to change in the next decade as well. For example:

Auto insurance is at a high risk of becoming irrelevant as we get autonomous vehicles. Rather than insuring drivers and vehicles, we may insure liability for road construction, autonomous driving software and true “act of god” incidents.
Home and some property classes will be more automated and premiums will be calculated based on sensor data feeds from IoT devices, based on things like risk of water leaks, how much traffic and activity there is in the house, and how many times doors and windows are left open or unlocked.
Workers compensation will evolve and become largely about proactive workplace safety rather than paying for losses post-accident. With IoT, augmented reality and other such developments, the number of workplace injury incidents will come down, as will losses. This will be offset by insurance companies spending more on safety promotion and monitoring, supporting healthy workplace practices and supporting return-to-work for injured workers.
Newer risk classes will emerge or grow, such as cyber insurance, drones, software/AI liability, terrorism (hopefully not), protection for alternative financial assets/transactions, and geopolitical risks.

Q: How will insurance products and services be distributed?

We envision significant elimination of the basic processes requiring human intervention and therefore a reduction in the headcount of insurance companies—perhaps down to about one-half to one-third of the current workforce.

Pasricha (Intellect SEEC): We expect most product classes to be distributed direct to consumer and the role of the intermediary changing to that of a risk assessor and consultant for more complex lines of business. In many cases, the current value chain (reinsurer to insurer, MGA to agent to consumer) will be significantly shorter, where major capacity providers may even directly offer protection to end customers. We also expect self-insurance trends to increase with emergence or strengthening of platforms for large enterprises to get coverages like excess-of-loss coverage on demand and without intermediation. Additionally, we anticipate a large increase in asset classes that will automatically be covered on a UBI (usage-based insurance) basis where a customer may not need to buy an annual policy. Instead, on every use [of the asset], a finite sum is auto-debited to the underwriting entity or an affiliated service provider.

Read more Future Insights by person