KEY TAKEAWAYS:

--AI, perhaps based on ChatGPT, can streamline customer support processes and remove much of the manual effort in data analysis, freeing time for analysts to do more important work.

--AI can also recommend predictive maintenance to clients and revolutionize quality control by detecting even subtle defects, reducing losses.

--While taking over certain types of work, AI will also lead to the emergence of novel roles such as AI ethics officers, AI trainers and data scientists.

----------

With the rapid advancement of artificial intelligence (AI), concerns about job displacement have become prevalent. However, it is essential to understand that AI, including language models like ChatGPT, has the potential to enhance and transform job roles rather than replace them. This article aims to dispel the notion that ChatGPT and AI are job killers and instead highlights how they can augment human capabilities and drive positive change in all industries across the scale.

The Evolution of AI and the Changing Job Landscape

AI technologies have made significant progress, with models like ChatGPT demonstrating remarkable language understanding and generation abilities. Instead of perceiving AI as a menace, we must embrace its potential to transform various industries, ushering in greater efficiency, enhanced customer experiences and the emergence of fresh employment prospects.

Streamlining Customer Support

AI-powered chatbots, such as those built on ChatGPT, can be invaluable in almost every industry for streamlining customer support processes. These chatbots can handle routine inquiries, provide in-depth information and facilitate various customer interactions. With these repetitive tasks automated, human agents can focus on complex customer interactions that require empathy and personalized attention. This focus improves overall customer satisfaction and enables agents to build stronger relationships with customers.

For instance, an insurance company's customer support team can use ChatGPT to respond instantly to common policy questions, allowing customers to receive immediate assistance without waiting for a human agent. This collaborative approach ensures efficient support delivery while freeing human agents to tackle more intricate customer issues.

Improving Data Analysis

Efficient data analysis is vital in various industries, and AI technologies such as ChatGPT can play a significant role. These advanced systems can analyze vast amounts of data, extract valuable insights and automate the initial evaluation. By reducing the need for manual data analysis tasks, AI streamlines the process and enables faster decision-making and actionable outcomes across different sectors.

In the case of an insurance company handling property damage claims, ChatGPT can analyze the submitted photos, assess the extent of damage and recommend an initial estimate based on historical data and predefined guidelines. Human adjusters can then review and validate these recommendations, ultimately expediting the claims settlement process and ensuring a seamless customer experience.

Predictive Maintenance

AI-powered systems like ChatGPT can be instrumental in various industries for implementing predictive maintenance strategies. By analyzing data from sensors, equipment logs and maintenance records, AI can identify patterns and anomalies that indicate potential equipment failures. This enables maintenance interventions, preventing costly breakdowns and maximizing operational efficiency across different sectors. 

For instance, an insurance underwriting team can use ChatGPT to evaluate potential policyholders' applications by considering various risk factors such as demographics, credit scores and driving records. ChatGPT can analyze the data, identify patterns and generate risk scores or recommendations for underwriters to consider during their evaluation. This collaborative approach improves risk assessment accuracy, efficiency and consistency, enabling underwriters to make well-informed decisions.

Quality Control and Defect Detection

AI systems, such as ChatGPT, can revolutionize quality control in various industries. By analyzing product data, sensor readings and historical performance metrics, AI can quickly identify deviations from expected standards and detect potential defects. This enables timely interventions, reduces waste and improves overall product quality, ensuring customer satisfaction across different sectors.

For example, an insurance company can deploy ChatGPT to analyze multiple claims submitted by the same policyholder within a short period. If the model detects patterns indicating potential fraudulent behavior, it can alert human investigators to conduct a thorough investigation. This collaboration between AI and human expertise strengthens fraud-detection measures, protects the integrity of insurance operations and reduces losses.

Creating Job Opportunities

Contrary to the fear of job displacement, AI implementation creates job opportunities. As AI technologies become more prevalent, the demand for professionals who can develop, deploy and manage these systems will rise. Insurance companies will require skilled individuals to train and fine-tune AI models, analyze data and ensure the ethical and responsible use of AI. Additionally, the integration of AI in insurance operations will lead to the emergence of novel roles such as AI ethics officers, AI trainers and data scientists, fostering innovation and driving industry growth.

Conclusion

Rather than fearing the impact of AI and language models like ChatGPT, we should embrace the transformative potential they bring. By automating repetitive tasks, enhancing decision-making and creating job opportunities, AI augments human capabilities and improves efficiency and customer experiences. As the insurance industry evolves, professionals need to adapt, upskill and collaborate with AI technologies to shape a future where humans and AI work together harmoniously to achieve greater success in the insurance landscape.

KEY TAKEAWAYS:

--Because they facilitate interactions, web portals can be used to gather data on customers and tailor products and services to them. Portals can provide a better customer experience and increase retention.

--Portals can also streamline business operations, saving time and money, while improving data security.

----------

With the global insurance market expected to grow 9% a year and reach $8.4 trillion in 2026 (according to Statista), competition is so fierce that insurance organizations need to stand out in every way they can.

One means that is too often overlooked is the development of a web portal. Unlike websites, which are primarily used to sell products or services, portals can simplify and speed communication and transactions between service providers and customers, also serving as a customer support tool, with a broader range of interactive elements. Portals usually require authentication and offer highly personalized content targeting specific groups. Moreover, a typical insurance portal provides automation and self-service capabilities, which helps manage an insurance business.

This article covers the five main benefits of web portals from a business perspective.

Increased customer retention

With the proper functionality, a web portal can help an insurance provider retain its customers continuously. For example, an insurer can send automatic reminders to keep customers engaged by notifying them of discounts and special offers.

In a more advanced scenario, an insurer can personalize these reminders to make them even more effective and engaging. Such personalized alerts can help insurers improve the quality of their service and sell more to existing customers, maximizing their lifetime value. Insurance providers can also notify customers when they should pay invoices, helping them avoid late payments while improving collections.

Outstanding customer experience

An insurance portal can also help insurers tailor their products to different audience segments and provide customers with self-service options demanded by modern customers.

According to the Future of CX: 2022 report by Freshworks, 39% of customers prefer self-service over any other channel, while 61% admire organizations that offer a mix of self-service and human touch.

Improved data security

The cost of cybercrime will reach $8 trillion this year and increase to $10.5 trillion by 2025, according to the 2022 Official Cybercrime Report by eSentire. Fortunately, by following specific cybersecurity practices, web portal developers can help insurance providers protect customer and corporate data from most cyber threats.

Data encryption

With data encryption, all information stored or processed by a web portal is coded and hidden from potential attackers. Only users with a corresponding key can access and view such encrypted data.

Role-based access control (RBAC)

By implementing the RBAC model, developers can limit access to web portal data and built-in functionality based on the user's role and position within an organization. With RBAC, an insurer can significantly reduce the risk of cyber-attacks, including those that use social engineering, while reducing the potential attack surface.

Data backup

Developers can back up insurance portal data using local or cloud storage. Even in the case of data leakage, it can be quickly recovered with a backup, helping an insurer prevent complete data loss and avoid unnecessary expenditures of time and money.

However, all the recommendations can only help if an insurance portal is developed correctly and securely. Therefore, developers must adhere to the latest cyber security practices and have validated software development certifications (such as ISO/IEC 27001 and ISO 9001).

Meanwhile, organizations should monitor their insurance portals and run regular security audits to ensure maximum security. If an insurer does not have enough resources for security maintenance, it is worth considering managed IT security services.

Optimized business performance

An insurance portal can make an insurance business more productive through automation and reduced manual labor. For instance:

Customer management

An insurer can manage its entire customer lifecycle and specific aspects, such as onboarding and retention.

Quote and policy management

Insurers can manage multiple policies in parallel and renew or cancel them as needed.

Underwriting management

Insurers can quickly assess the insurance risks of their applicants and calculate the amount of coverage for each of these risks.

Inquiry management

Agents can prioritize inquiries, manage them and track customer payment history.

Claim management

Insurers can streamline and expedite claim registration and processing by implementing the right tools.

Document management

With robust functionality, insurers can quickly search for necessary files and data, share them and promptly generate documents (such as policies, claims or invoices) based on predefined templates.

These are only some features to integrate into a web portal to improve insurance business performance. For example, depending on its business requirements, an insurer can instruct developers to adopt tools for accounting, reinsurance and automated marketing.

More advanced analytics

Typically, insurance portals store and collect large amounts of data. Insurers should empower their web portals with in-built analytics features to reap the benefit of all this data.

For example, an insurer can track sales and transaction data and then apply machine learning models to predict demand for specific products or services. Or, an insurer can track usage metrics, such as login rate, session duration and bounce rate, to make the insurance portal more efficient and user-friendly.

Final thoughts

The insurance industry is growing, but this growth has a downside. Each day, service providers enter the insurance market, leading to more intense competition. Adopting digital technology is a natural step for and standing out. 

If developed properly, an insurance portal can provide outstanding customer experience, enhance customer retention and streamline business performance. In addition, a robust web portal will guarantee the safety of customer and corporate data.

KEY TAKEAWAY:

--As insurers replace board members, they are often looking these days for executives from other industries. Those executives can provide valuable, fresh perspectives on technology, customer experience and more but can benefit from focused education on the unique issues of insurance company governance.

----------

Recently, I was one of three people who retired from the board of an insurance company I had long and proudly served, because we had reached the company’s mandatory retirement age for directors. I am just one of hundreds of Baby Boomers who have aged out of their insurance company board seats in recent years. The enormous size of the Boomer age cohort has produced a flood of board retirements at insurers, and this trend will continue for several more years as the youngest Boomers move into their 70s. As a result, the industry is experiencing a brain drain of highly experienced board members. 

New Ideas and Perspectives

This age wave has, in turn, resulted in a surge in new board members at insurance companies. New people, new ideas, fresh perspectives. All good. But this changing of the guard is different than in years past. Insurers are seeking to build more diverse boards than those traditionally dominated by white male former CEOs. Armed with solid empirical support that diverse boards produce better governance, companies are seeking diversity not only in gender and race but in age, skill sets and perspectives, as well. 

One of the diversity characteristics that insurers increasingly seek in board candidates is experience in other industries, to enhance insurers’ understanding of technology, the consumer experience, product and distribution channel preferences and other non-traditional insurance areas of focus about the changing world around us.

Adding a chief technology officer, an economist, a consumer products executive or a marketing executive from another financial services sector, for instance, can bring valuable new perspectives to an insurer’s board. I’ve seen this happen many times, and it can energize a board to have fresh eyes look at their concerns.

The board member’s value to the company can be enhanced and accelerated, though, by having them receive focused education on the unique issues of insurance company governance.

Understanding the Business of Insurance

Insurance company directors don’t need to become experts in insurance accounting or taxation, develop the knowledge to settle a workers’ compensation injury claim or underwrite a cyber risk policy. But there are issues of insurance company governance and management oversight that are different from other industries and that they do need to understand. If learned from experts in the field, this education can make a board member more effective in a shorter time. The objective is, after all, about governance capability, not management skills.

Drawing on its unique membership of insurance thought leaders from around the world, the International Insurance Society has developed a board education program tailored for existing and prospective insurance company board members. From its insurance executive, risk management academic, consultant, accountant, actuary, asset manager and regulator members globally, the IIS has assembled ideas, experiences and perspectives that constitute the best practices of insurance company governance. 

The program launched in 2022 and was led by a teaching panel of industry experts with complementary skills and insurance board experiences. Modules include how various insurance functions and issues can be best understood from the board perspective, recurring areas of insurance oversight challenges and lessons learned regarding board successes and failures in dealing with major issues. Participants have included both life and property/casualty insurance directors from companies ranging from giant multinationals to regional mutuals and were enthusiastic in their reviews of the two-day program. 

The participants were all chosen for a board position by their companies because of their notable success in a wide range of career fields but felt strongly that the insurance-specific curriculum made them better able to do their job as board members of an insurer. They benefited not only from exchanging ideas with the faculty panel but with each other, as well.

We heard from the attendees about the value they felt they gained and subsequently heard from their sponsoring companies that the program was a good investment for them, too: The program accelerated the timeline of getting board members up to speed and fully able to understand the critical issues facing the company. 

What Attendees Say

“The faculty were able to provide real life scenarios that demonstrated the responsibilities of governance as a board member versus responsibilities as executives in insurance companies. The faculty did a good job pointing out the unique financial considerations that board members should monitor. It was also very helpful to interact with the other attendees, all of whom had different experiences and were most willing to share,” said Dorie Culp, member of the board for Utica National Insurance Group.

Perry Hines, independent member of the board of Horace Mann Educators, said, “The program was highly interactive, and it was great to interact with industry legends facilitating the program. It was an impactful two days that really solidified an understanding of insurance and governance.”

Gisselle Acevedo, psychotherapist, educator and member of the board of governors for Farmers Insurance recommends the program for insurance board members with all levels of industry experience. She said, "I was so impressed by the training that I ensured other board members also attended. In addition, the material that was covered was relevant and has continued to help me. There were in-depth dives that, in all the training I have had throughout my many years, I had not received before. It is essential for any insurance board member to attend, no matter how seasoned they are."

Explore the Experience

Please contact Jason Terrell at jterrell@internationalinsurance.org at the IIS to reserve a spot for yourself or to nominate an existing or prospective director of your company, for the next IIS Board Education program. You and your company will both benefit.

Agents and brokers have been told for years now that they need to enable clients to do as much self-service as possible. That way, clients can access their accounts however they want and whenever they want -- every hour of the day, every day of the year -- without having to reach their agent. 

But that recommendation runs up against reality in this month's interview, with Jason Keck, founder and CEO of Broker Buddha. And he would know. His company produces software that automates swaths of work for agents and brokers -- mostly related to the intake of information so they can submit applications for quotes -- so he sees automation up close all the time.  

He says that, sure, some interactions should be turned over to self-service. "Some self-service things can be automated, no question," he says. "I need to get my policy documents. I need to get my ID card. Tools for those sorts of things are great.... The payment space is definitely an opportunity. You're seeing a lot of really good adoption there."  

But, he adds, "As soon as you head into making a change to your policy, it's harder, right? Who's your policy with? Will there need to be a premium change?

"Policy changes and endorsements require a lot of hand holding and workflow in the back office. It's hard to build a tool that connects the insured to the broker and to the carrier and facilitates that end-to-end workflow. You don’t just have to build a product for one user group; you have to put together a product that serves three user groups seamlessly and that everybody adopts. That’s really hard."

So, while I still favor as much self-service as is possible, I don't favor MORE than is possible, and it sounds like a considerable amount of work still can't be automated.  
 
For now.... 

Cheers,
Paul

P.S. Here are the six articles I'd like to highlight this month for agents and brokers:

WHY AUTOMATION IS SO IMPORTANT

Agencies, brokers and carriers spend an inordinate amount of time and money on back-end processes that detract from core objectives.

GAMIFICATION COMES TO LIFE (INSURANCE)

Finance games have been shown to greatly help with life insurance sales -- and digital, large-scale versions are becoming available.

'INTELLIGENT INGESTION': TIME TO TRULY GO DIGITAL

The industry kids itself about having gone paperless. In fact, we still use the same processes we used in the 17th century. It's time for a change.

CHANGING EXPECTATIONS ON MOBILE PAYMENTS

41% of millennials with insurance purchased it with their mobile device, and other generations are moving in that direction, too.

WE MUST BE DIVERSE BY DESIGN

We crave diversity because we know it serves all people with speed, intelligence and transparency and delivers better business outcomes.

STOP CALLING MY DAUGHTER!

When my daughter recently looked online for information about health insurance, companies showed how NOT to sell to millennials.

ITL:

You’ve been a force in software for agents and brokers for some time now. So, I’ll start by asking, generally, what you see as the state of agencies at the moment and where you think they’re headed. Then I’ll poke a bit.

Jason Keck:

One thing I’ve learned is that if you’ve met one agency, you’ve met one agency.  You can’t generalize insurance agents or agencies, and it’s hard to categorize them, as well.

What I have noticed, though, is that larger agencies have institutionalized technology rollout and adoption capabilities. Those agencies are having the most success adopting tools outside of the AMS [agency management system]. In particular with the super-regionals, you have very high-quality, institutionalized adoption. They're very thoughtful and methodical.

As you get into the smaller independent agency channel, it's really a function of the workforce. I don't mean the principals. I mean the actual people with the desks. In some places, you have a whole group of people who are generally younger and more digitally native, and they're very capable with technology. In some places, you don’t. There isn’t a one-size-fits-all answer.

ITL:

If you're running an agency, let’s say an independent agency, how would you think about becoming more efficient?

Keck:

I would first try to understand where my team is spending most of their time, and I’d start by looking at the org structure for the agency. Some agencies are organized by line of business, some by size of business, some by class of business, some by workflow (for instance, new, renewal, service, marketing). Then you have to jump in, searching for where people are spending their time.

Newer agencies are going to be a lot more focused on new business. More established businesses are going to be a lot more focused on renewals.

Once you do the segmentation, you can decide where you’re going to optimize and drive efficiencies. When you figure that out, you ask, what are some of the tools out there to make me better?

ITL:

What are some of the tools out there – other than, Jason’s phone number is…?

Keck:

The payment space is definitely an opportunity. You're seeing a lot of really good adoption there.

But the tools for service are really poor. You have some policy portal and service capabilities, but there’s an opportunity to do a lot better.

The same is true on the commercial rating side. I’m not hearing about anything that’s flying off the shelf. There’s no, “Man, you gotta do this or else.”

ITL:

I'm really curious about service. We publish a lot about it, and the need for improvement seems obvious. Why aren’t the tools better?

Keck:

I haven't seen any good new service technologies in my five years at Broker Buddha. Why? Because CSRs [customer service representatives] are relatively low-cost, and there's a decent amount of mental capacity that has to go into service. Policy changes and endorsements require a lot of hand holding and workflow in the back office. It's hard to build a tool that connects the insured to the broker and to the carrier and facilitates that end-to-end workflow. You don’t just have to build a product for one user group; you have to put together a product that serves three user groups seamlessly and that everybody adopts. That’s really hard.

ITL:

People talk a lot about the need for self-service. It sounds like you're saying self-service may be harder than people think.

Keck:

Some self-service things can be automated, no question. I need to get my policy documents. I need to get my ID card. Tools for those sorts of things are great. But as soon as you head into making a change to your policy, it's harder, right? Who's your policy with? Will there need to be a premium change?

ITL:

Let’s turn to Broker Buddha. What sorts of problems are you guys solving, and where do you think you can go from here?

Keck:

We've been incredibly successful at solving the submission intake problem, particularly around specialty lines. That's our bread and butter, both for new business and renewal. Filling out a PDF sucks whether you’re doing it the first time or having to do it the second time. It's a super, super frustrating experience.

Where are we going from here? The small commercial intake process is an area of white space that nobody's really done a good job of solving for. You're trying to collect information to populate an Acord form. But agents don't know how to complete the form or which questions are necessary to actually get a quote. They don't even know how to explain the questions to their clients! They don't know.

And you also have to populate multiple forms. As an agent, you end up needing to enter the same information on different forms. So, we’ve built a really elegant solution that dynamically loads the forms you need, hides the questions that are irrelevant, defaults answers to the most common responses, uses conditional logic to reveal important questions and generates the PDFs you need when you’re done.  Plus, we integrate eSignature into the experience to tie it all together.   

That's number one. Number two is something that we overlooked early on but that I'm getting really excited about. Most agents are really familiar with quoting in carrier portals, but the portals are really frustrating. There's a material problem around collecting the information needed to enter into a carrier portal because agents don’t know the questions the carrier asks when they’re talking to their customer. This means that when the agent tries to get the quote, they may not have all the information they need. We’re replicating carrier portal questions to support the intake process so the agent never has to go back to the client a second time for more information.

ITL:

Despite all the talk about disintermediation, it’s clear that agents have won. Where do you see them going if you project out a few years?

Keck:

If you look at something like cyber, you see that the macro landscape is having a material impact on insurance products, coverages and pricing. The agent’s role, in many ways, is to understand the changes and to help clients reduce their risk. That's a full-time job, staying up to speed.

What is and isn't covered on these policies becomes very important, and how much you have to pay for specific endorsements makes a big impact. And that's nuanced stuff.

ITL:

That's pretty much what I wanted to hit. Is there anything I didn't ask about that I should have asked about?

Keck:

Change management needs to be a big focus. We started out with a very robust product for small agencies, then moved up-market to larger agencies. What we're doing now is actually stripping down a lot of what we offer so smaller agencies can more easily adopt it. The key is simplicity, which makes change easier, even for less digitally native agencies.

We're still building out all kinds of workflow tools for larger agencies in complex situations.  At the same time, we’re rolling out a simpler solution for smaller agencies that won’t overwhelm them. Solutions have to be turnkey for smaller agencies.

ITL:

Thanks, Jason.

KEY TAKEAWAYS:

--The typical auto insurance model, which has been around for a century, notes the base price of a new vehicle and assumes roughly 20% depreciation each year. But the pandemic has uncoupled that model from reality.

--In recent years, scarcity of vehicles and other issues have resulted in a surge in used car values – sometimes going above their price new, often holding value instead of plunging and generally not depreciating as in “normal” times.

--Every pricing system, every actuarial triangle and any predictive model using the last 20 years or more of historical data is ignorant of this kind of volatility and unprepared to say what should happen next. It's time for insurers to change the way they look at vehicle values. 

----------

If you were to rent a car, you would expect that a more expensive vehicle would cost you more. You would also expect that, if you damaged that more expensive vehicle, it might cost more to repair. So, you would not be surprised to see a higher-cost insurance option associated with more expensive (and perhaps more powerful or luxurious) versions.

Well, your regular insurance program for your vehicle is not working quite like that – at least not yet.

While the base price of any vehicle can move your insurance rate north, as you add on optional equipment, most current insurance pricing systems don’t take those new features into account. Existing pricing models ignore anything beyond the base price new from an MSRP on the window sticker. Maybe more surprising, that is the last time they think about the value of your car.

Why? Let’s take a step back in time.

The very first auto insurance policy is attributed to Travelers in 1897 for a risk in Ohio. Other auto insurance companies, such as State Farm and USAA, recently celebrated 100-year anniversaries for insuring consumer vehicles. While it’s exciting to see brands across the industry celebrate such a marker, we also just weeks ago saw the very first loss in an annual report in 100 years for one insurance company – something we think can largely be attributed to missing the changing conditions on the value of a used vehicle.

Over the last century of auto insurance, paper policies managed in mailboxes and inbox bins have been increasingly replaced by digital policies on smartphone apps in the cloud, but many legacy processes have not kept up with the times. Two iconic pieces of auto insurance history come to mind – the use of base price new MSRP to set a value in a “set it and forget it” fashion and the use of a one-size-fits-all factor curve for setting a relativity-based insurance value for a vehicle over time (across the nation or within a state).

Why should these processes be left in the past?

See also: Deteriorating State of Car Insurance

We will save the window sticker story of base price versus total price and the gap between the two for another day. Today, we want to spend a moment exploring active valuation methods versus static and embedded methods for vehicles.

Model risk management disciplinarians may not have gotten around to vehicle valuation for pricing yet – the use of base price new and no further detailed tracking of the asset is deep in tradition and only noticeable in filings and rating algorithms, so hidden in plain sight. The current displacement in valuation trends, however, has completely uncoupled the traditional approach from today’s reality.

Let's explore.

Model Risk Matters - Active vs. Static Methods

One of the more longstanding traditional presumptions since the time before computers were introduced, back when data was manipulated with slide rules and look up tables, is that vehicles always get cheaper with age. That is not a shocking thought, but in recent volatile car valuation times, this embedded thinking has turned into a systemic risk (model risk). Let’s look at a specific case and then the general case.  

The Toyota 2017 Rav4 model year value trend: A specific example in these pandemic era times

Prior to April 2020, the market for a used 2017 Rav4 followed a traditional rule of thumb: straight-line depreciation of about 20% a year. That meant that, in five years, it would have lost 2/3 of its original price. $30k-ish moves to $10k-ish across 60 months. (See the green line in the Rav4 exhibit.)

While other things that consumers insure, like houses, rings, art, etc., all seem to cost more over time – sometimes intrinsically and sometimes due to inflation – the cost for cars normally is presumed to trend downward as they wear out their mechanical function with usage. Nowadays, however, are not normal.

In recent years, scarcity of vehicles and other issues have resulted in a surge in used car values – sometimes going above their price new, often holding value instead of plunging and generally not depreciating as in “normal” times. (See the gray line in the Rav4 exhibit.) Values can also vary from type of vehicle to brand of vehicle and perhaps to the trim level and optional equipment for popular configurations – not a one-size-fits-all situation.

The J.D. Power Used Vehicle Price Index (UVPI), constructed to measure relative market valuation for vehicles less than eight years old (a classic plan for targeting under 100,000 miles on the odometer), indicates that, since the turn of the millennium in the year 2000 until the beginning of the pandemic in 2020, there were relatively “calm seas” in the used car price space.

J.D. Power Used Vehicle Price Index: Trend from January 2000 - April 2023

We observe a notable downturn during the global financial meltdown in late 2008 (and follow-on recession), with a new plateau in pricing slightly higher for post-recession than pre-recession. There were many insurance pricing increases put into effect to adjust to that new plateau of value of vehicles about a decade ago, but otherwise we see a relatively uneventful 20 years of valuation on an indexed basis.

That is until the pandemic.

What we see today indicates that a sea change like never before has hit the shore of used vehicle values. Every pricing system, every actuarial triangle and any predictive model using the last 20 years or more of historical data is ignorant of this kind of volatility and unprepared to say what should happen next.

See also: Buckle Up for Telematics 2.0

It's time for insurers to change the way they look at vehicle values.

Today, it seems every auto insurance company is raising rates across the board on a more frequent basis and in higher amounts than at any time in our lives. Even with the current rising tide, however, we have not seen enough increase to match the change in the most valuable parts of the car fleet, which are up almost double (from the 1-teens to the 2-teens) in the most recent months of the used vehicle price index (UVPI).  

While it is impossible to know what will happen next, it seems the insurance industry can comfortably assume it needs to really pay attention to the value of insured vehicles in a different way than they have so far, to be around for the next 100 years.

This article originally appeared on the J.D. Power site here.

While I've been a proponent of embedded insurance, as have many others, an article published recently should cause some pondering. The article argues that embedding insurance is really just a way to pester customers to buy coverage they don't need.

"You aren't going to take a chance on dinging that rental car, are you?"

"Yes, we just recommended that laptop you're purchasing, but, you know, that model breaks down a lot, so we highly recommend buying a warranty."

"Your trip isn't covered. Are you sure you don't need travel insurance? Are you really sure? Really? We'll give you one more chance...."

The article took me back to the arguments I heard in the early days of mobile phones, in the early to mid-'90s, when carriers talked about how great it would be for, say, coffee shops to know that a customer was walking by and to be able to ping their phones with a coupon for a latte. So many people rejected that idea as nagging, even stalking, that carriers backed off.

I still think embedded insurance, done right, can make the purchasing process more efficient and take out costs while expanding the reach of certain types of insurance, both existing types and potentially new ones. But it's worth looking at the counterarguments, if only to make sure that embedded insurance is, in fact, done right.

Ian Gutterman, founder and CEO of Informed Insurance, makes the argument against embedded insurance in two parts, first laying out why he believes it's bad for consumers and then explaining why he thinks it's even bad for insurers.

He writes:

"Almost every type of insurance being offered as 'embedded' could easily be bought separately. We are not talking about creating new types of insurance for novel risks. It is mostly auto, pet, comp and warranty.

"The main reason they are being repackaged as 'embedded' is because the insurers realized they are bad at selling them on their own. The hope is by using a middleman they can catch you in a weak moment where you might buy pet insurance from the pet store that just sold you the puppy.

"It’s emotional manipulation rather than truly addressing a need."

Gutterman notes that proponents of embedded insurance often argue they're just nudging people toward coverage they need but says, "more often than not, the motivation is what is good for the corporate interests rather than the consumer."

Despite that emphasis on corporate interests, he thinks that embedded insurance is bad for insurers, too, at least in the medium to long term, because it makes their insurance a commodity that can be swapped out for someone else's at any moment. 

Gutterman writes: 

"When you choose to be an embedded insurer, you are... a nameless, faceless entity. The customer doesn’t care who you are. They attribute all the value they perceive from the transaction to the retail brand that you partnered with.... [And] more often than not, the retail partner couldn't care less if another insurer stood in your shoes....

"You don’t have to be a game theory expert to understand how this will end. Distribution partners will demand bigger and bigger pieces of the pie. If you don’t cut them another slice, they will drop you for another partner."

He also warns that embedding the insurance could lead to adverse selection:

"The partner will want to sell your product to every customer since that is incremental fee revenue for them, even the ones with a high likelihood of a claim or willingness to commit fraud. Unless you demand they receive a healthy chunk of their payment as profit-based contingent commissions, they won’t care about your loss ratio."

Gutterman says embedded insurance will only work if it provides three things:

• "Better coverage than available in traditional channels
• Lower price than in alternative channels
• Easier buying process with clear explanation of why it’s a good price and offers better coverage"

That's a pretty daunting set of arguments.

But before you give up hope, I suggest looking at a piece that Chris Bassett, a senior director with Capgemini Invent, published on ITL in January. He says embedded insurance won't fulfill its potential if the coverages being sold are only the sorts of point-of-sale products that Gutterman disparages. But Bassett has high hopes if embedded insurance moves to "point-of-design."

He writes: 

"The opportunity lies in exploring the affinity between the insurance solution and the non-insurance product and how we can build unity in a singular value proposition. Doing so requires introducing insurance at the point-of-design, rather than as a bolt-on at the point-of-sale."

As potential examples, he suggests:

• "Connected home solutions with advanced safety and maintenance features built into the property that detect and automatically manage risk (turning off appliances or mains supplies and automatically contacting emergency services)
• Personal vehicles designed to support wellness-focused internal recreational spaces as an extension of a bundled life, health and vehicle insurance package
• Sports equipment (shoes and clothing) with sensors woven into them that monitor performance and health as an extension of a goal-directed training and wellness package that includes personalized life and health insurance risk mitigation and protection, and that connects to a branded community of similar users."

For me, Gutterman describes a scary series of potential land mines, while Bassett lays out a way to start navigating through them. But I encourage you to explore further, both with the full pieces I've linked to here and, perhaps, with some of the other smart pieces on Gutterman's blog.

Cheers,

Paul

KEY TAKEAWAYS:

--The bottleneck is the time it takes to manually review if the presented information is correct. That can take days, leading to delays in policy issuance.

--Software powered by AI, blending multiple layers of technology, can seamlessly conduct identity verification in real time.

--Better verification can help reduce fraud, which cost consumers and businesses more than $300 billion in 2022 just in the U.S., and can help deal with complex compliance requirements.

----------

Despite the industry's stodgy reputation, insurance is often ahead of the curve, embracing new technologies that improve efficiency, enhance customer experience and increase profitability. Artificial intelligence (AI) and machine learning (ML) have been used to analyze vast amounts of data that help identify patterns, trends and insights that can help insurers better understand risks, predict claims and personalize their services. Blockchain has been used to improve the efficiency of claims processing, reduce fraud and enhance transparency. Mobile apps, such as telematics, and wearables have allowed insurers to collect real-time data on customer behavior and offer personalized pricing and services. Robotic process automation (RPA) and chatbots help automate routine tasks and improve customer service. 

By contrast, the insurance industry has been hamstrung by outdated identity verification practices. Currently, verifying a customer’s identity consumes unnecessary time, labor and capital.

The bottleneck is the time it takes to manually review if the presented information is correct—that can take days, leading to delays in policy issuance and a poor customer experience. Insurance carriers, brokers, regulators and customers can all gain an advantage from a secure and expedited transfer of personal information. 

Existing methods do little to combat the fraud plaguing the insurance industry, which the Coalition Against Insurance Fraud estimates cost businesses and consumers more than $300 billion in 2022. During the same timeframe, the FBI found that fraud was costing the average American family between $400 and $700 a year in premiums. Life insurance topped the list of sectors plagued by fraud—totaling $74.7 billion in lost revenue. Next came Medicare, clocking in at $60 billion, then property & casualty insurance, with $45 billion in lost revenue. 

Despite these discouraging statistics, recent advances in cutting-edge technology show incredible promise for the detection and prevention of insurance fraud. A survey of 100 insurers across a broad spectrum of the industry by the Coalition Against Insurance Fraud found a steady progression in the adoption of detection of claims fraud technology, from 73% in 2014 to 96% in 2021. The same pattern was observed in underwriting or point-of-sale fraud/rate evasion technology, where each year showed an incremental increase in adoption, from 28% in 2014 to 65% in 2021. 

Advanced identity verification technology was the outlier. It was not observed in all previous studies (in 2014, 2016 and 2018), because advanced identity verification technology is new and groundbreaking—pioneered in just the past couple of years. Four in 10 respondents said in the 2022 study said they had adopted such technology.

This software is powered by AI, blending multiple layers of technology to seamlessly conduct identity verification in real time. Anyone who has integrated a next-generation provider of identity verification has seen the benefits in both their onboarding numbers and their operational costs, and the percentage of those employing these solutions is predicted to drastically increase over the next three to five years. 

Modernizing identity verification practices isn’t just about saving time and labor and enhancing the customer experience—it's key from a regulatory compliance angle. The Gramm-Leach-Bliley Act (GLBA) protects the privacy of customer information and requires safeguards to prevent unauthorized access to that information. Enhanced identity verification practices provide a two-way improvement in compliance to GLBA—easier compliance related to data security, as well as secure provisions related to processing customer identity information. The Fair Credit Reporting Act (FCRA) regulates the collection and use of credit information, including collection done in the context of insurance underwriting. Modernized identity verification practices can help insurance companies comply with the FCRA by providing more accurate and complete customer identity information, which can be used to make more informed underwriting decisions.

There is a big price to pay both financially and reputationally when robust measures are not in place. Anthem was fined $16 million for a data breach in 2015 that exposed the personal information of nearly 79 million customers. In the aftermath of the breach, Anthem was criticized for not doing enough to prevent the attack and for not detecting the breach sooner. 

See also: Applying Cyber Lessons to Regulating AI

Britain is raising the bar for fraud prevention regulations by not only penalizing companies found guilty of committing fraud but also the people who fail to adequately prevent it. The Economic Crime and Corporate Transparency Bill, which is in the process of being passed, enables companies to be fined without limit, even if they were not aware of any fraudulent activities taking place within their organization. The proposed legislation is designed to create a culture of aggressive fraud prevention.

The American insurance industry has received plenty of bad publicity during the last decade—users claim the process of getting insurance is time-consuming, confusing and overly bureaucratic. This dissatisfaction starts during the onboarding phase, which requires users to remember unique passwords and lengthy security questions. Biometric technologies like liveness testing and facial matching are a faster alternative to traditional verification methods by enabling customers to verify their identities instantly on a device. Coupling this with document verification to ensure what’s submitted for identity verification purposes has not been forged or tampered, and with address verification—where the individual’s name and address are extracted from the onboarding information and verified against multiple databases—can expedite claims decisions by instantly checking passports and driver's licenses and removing the need for submitting a proof of address document. What was once a tedious and friction-filled onboarding process can become intuitive and user-friendly—boosting brand loyalty and customer satisfaction.

The future of technology adoption in the insurance industry looks bright—by incorporating advances that can help prevent fraud, comply with regulations and enhance the customer experience. Modernizing identity verification practices means ushering in a new age of security and profitability—benefiting customers, regulators and insurance providers alike.

KEY TAKEAWAYS:

--While traditional new business and renewal processes can be time-consuming and result in missed opportunities, AI can quickly generate alternate plan designs.

--During high-load periods, the volume of quotes requiring underwriter review can slow processes. AI can assist by pulling together data, even from previously inaccessible sources like handwritten notes.

--AI-powered chatbots can provide immediate assistance and answers to customer queries, providing a better customer experience. These tools can be trained to learn an insurance company’s products, policies and general “language,” helping customers fully understand their benefits plan.

---------- 

Artificial intelligence (AI) has the potential to revolutionize the group insurance industry, particularly during peak business periods, when human resources are stretched thin. 

By streamlining quoting, optimizing resources and automating manual tasks, AI can help increase group insurance sales, boost profitability and improve the customer experience. As a result, 74% of insurance executives plan to increase their investments in AI. Insurers that seize this opportunity early will have a critical advantage. 

In this article, we'll explore the various ways in which group insurance providers can leverage AI to maximize sales and remain competitive. 

Recommending Alternate Plan Designs with AI 

Traditional new business and renewal processes can be time-consuming and result in missed opportunities. It’s helpful for sales and underwriting teams to have data-backed reference points when designing plans for clients. 

Using predictive analytics, insurers can quickly generate alternate plan designs using one of their most valuable assets: their data. 

Here’s how it works: Using an AI recommendation engine, carriers can use their historical sales data to identify the most successful sold plan designs for similar clients. Recommendation engines, like those offered by the Majesco Global IQX Sales & Underwriting Workbench, will fetch data for similar groups (size, geography, industry) and suggest an alternate plan design that can be presented alongside the handpicked plan.  

See also: The Importance of Explainable AI

This acts as a powerful benchmarking tool while circumventing the data limitations in group insurance because it uses the carrier’s own data. 

AI can also be trained to make suggestions for upselling and cross-selling opportunities on optional worksite products, improving the plan member experience while unlocking revenue opportunities. Some factors it might consider to recommend products may include: age, gender, activity level and job role. 

These tools not only help insurers tap valuable existing sources of revenue but also reduce quote turnaround time, providing a better customer experience. 

Optimize Underwriting Resources 

During high-load periods, the volume of quotes requiring underwriter review can slow processes due to an inefficient allocation of human resources.  

Underwriters have varying levels of expertise across different products and quote complexities. Additionally, managers must take their existing workload into account. As a result, quotes might be sitting for days or even weeks before they can be turned around.  

AI can assist underwriting managers in suggesting the most effective distribution of quotes across the underwriting team, taking into account an individual underwriter's current capacity, their expertise and their performance history.  

Additionally, AI can prioritize quotes with the highest chance of closing based on past successes. By identifying these resource efficiencies, insurers can write more business in less time while improving close ratios. 

Automate Manual Tasks, Process RFPs Faster 

The influx of requests for proposal (RFPs) can produce unwanted friction and increase quote turnaround time.  

To assemble a group benefits quote, sales and underwriting teams need to collect and process several key pieces of information from different sources.  

These may include past plan booklets, policy summaries, employee census files, prior claims experience and general client information (industry, size, location).  

Here’s the problem: These data often come in different formats, and they are often not delivered in the form of structured data. In other words, the data cannot be automatically ingested by digital tools. As a result, underwriting teams lose countless hours to rekeying information into their internal systems. 

This Is Where AI Comes in.  

AI systems can extract information from broker emails, attachments and RFP source files using techniques like optical character recognition (OCR) to identify and process unstructured data trapped in PDFs, images, the “back of a napkin,” and other formats.  

Natural language processing (NLP) can be used to train an AI system to read an RFP booklet, learn carrier-specific terms and abbreviations and generate a quote based on the information detected within the source document. These automated processes not only save time but also reduce errors, providing a better customer experience. 

Once the raw data has been captured, it can be converted to structured data in the insurance company’s format and sent directly to their quoting portal or underwriting workbench.  

AI leveraged in this way can drastically cut the time it takes to assemble a group benefits quote, contributes to improved close ratios and fosters better relationships with your distribution partners. 

AI Chatbots to Drive New Revenue Opportunities 

Gone are the days of clunky, frustrating chatbots that pop up when you least want them. The release of ChatGPT in the fall of 2022 showed the true promise of generative AI. 

By using AI-powered chatbots, group insurers can provide immediate assistance and answers to customer queries, providing a better customer experience. These tools can be trained to learn an insurance company’s products, policies and general “language,” helping customers fully understand their benefits plan. 

Chatbots can also be used to identify potential upselling and cross-selling opportunities for voluntary benefits, improving overall sales and profitability. By integrating with health data providers (e.g., Fitbit) and behavioral data tools, this functionality can become even more personalized, tailored to the individual’s unique situation and their goals. 

While agents and brokers can't offer personalized advice to thousands of individual insureds with varying needs and levels of knowledge, technology can. Digital assistants and chatbots can provide valuable guidance to plan members, answering questions, guiding them through the enrollment process and recommending new products and services based on their health and demographic information. 

The Best Digital Assistants  

Sun Life's AI-enabled "digital coach," Ella, encourages plan members to take action on important deadlines, cost-savings and product offerings. Ella's nudges resulted in nearly two million additional interactions from plan members early in the pandemic from January to September 2020. 

Sun Life's investment in virtual coaching paid off. Thanks to Ella's interventions at the right time to the right user, the company experienced an 83% increase in additional coverage purchased compared with the previous year. This success demonstrates the power of digital assistants in providing personalized education and guidance to insurance customers. 

See also: Beyond the Digital Transformation Hype

Embracing the Potential of AI in Group and Voluntary Benefits 

The future of insurance includes AI, which is poised to revolutionize the group insurance industry, providing significant benefits to both insurers and customers.  

By streamlining quoting, optimizing resources, automating manual tasks and improving customer service, insurers can increase group insurance sales, improve profitability and provide a better customer experience.  

KEY TAKEAWAYS:

--Developing AI models with representative and unbiased data leads to increased accuracy and fairness in predicting outcomes and making decisions, resulting in more effective products that meet the needs of a broader set of users.

--But there are challenges, including that data that accurately represents the population and is not biased may not be available.

----------

Companies often err when it comes to getting feedback on their latest AI. As iterations progress, a team may discover that the algorithms and use cases are enabling misinformation or causing harmful outcomes for customers. At this point, even if the team retracts the product, customers will demand to know why harmful consequences weren’t found through testing before the product was released.

It is a scenario that puts the reputations of both you and your customers at stake.

The following guidance can help you assess where to adjust your product development and design thinking to make ethical AI an enabler of awesome products your customers will trust and love.

The Difference Between Ethical AI and Responsible AI

Although often used interchangeably, "ethical AI" and "responsible AI" are different. Because this post is focused on AI ethics and product development, it’s important to explain the difference between the two terms.

Ethical AI includes the principles and values that direct the creation and use of AI. It underscores that AI systems are developed and implemented in a way that coincides with ethical considerations such as accountability, transparency and impact, with people as the focus. Ethical AI tries to ensure that AI is built and used with justice, even-handedness and deference to human rights.

Responsible AI encompasses the measures and practices you’ve implemented to manage and plan for ethical use, in addition to aspects such as safety, security, accuracy and compliance. These practices include maintaining data quality, creating transparent and explicable AI systems, conducting frequent audits and risk assessments and establishing governance frameworks for AI.

It is important to have a responsible AI approach to ensure that ethical AI principles are effectively put into practice.

See also: The Rise of AI: a Double-Edged Sword

Ethical AI Principles in Product Design and Development

Product teams can maximize the potential of AI and enhance the effectiveness of their products by adhering to ethical AI principles. Ethical AI also promotes innovation in product development.

Here are some examples of where you should be looking in your design and quality checks for AI reviews:

• Developing AI models with representative and unbiased data leads to increased accuracy and fairness in predicting outcomes and making decisions, resulting in more effective products that meet the needs of a broader set of users.
• Incorporating ethical AI practices into the development of AI models increases transparency and explainability, improving user trust and driving more use of products perceived as fair and understandable.
• AI can automate tasks and processes, resulting in increased efficiency and reduced workload for users. But there are implications to consider about what tasks are being optimized and why. Adhering to ethical AI principles, product teams can create AI models that are optimized for reducing mundane tasks so workers can take on higher-value tasks that sustain future growth for themselves and the company.
• Ethical AI principles offer product teams the chance to explore novel opportunities and assess use cases for AI. By crafting AI models that are transparent, explainable and fair, product teams can demonstrate the value of their AI before it affects customers and society.

Adhering to ethical AI principles during development allows for the creation of AI models that align with core societal values and fulfill business objectives. The effort to improve product accuracy, effectiveness and user-friendliness for all stakeholders within an ethical framework enables product teams to leverage the potential of AI fully.

If it sounds like more stakeholders in the development process, such as user experience, data engineering, risk management and even sales, might be affected by ethical considerations when developing AI, your hunch is correct. Cross-team visibility will become essential to upholding both AI and corporate ethics.

Let’s explore the challenges.

Challenges for Adding Ethical AI Reviews to Products

Incorporating ethical AI principles into product development is essential for responsible and trustworthy AI applications. However, the following challenges and objections might arise during the stages of the process:

• Data that accurately represents the population and is not biased may not be available. Biased data can cause discriminatory and unjust outcomes when AI models perpetuate or amplify existing biases.
• Transparency is key to ethical AI practices, but achieving alignment across teams can be tough. Without designing for interpretability, AI models will lack transparency, which can hinder understanding of decision-making processes when issues arise and time to correct model behavior is critical.
• Likewise, a lack of transparency combined with disagreement on ethical policies can slow development. Early warning signs occur when stakeholders feel ethical principles are an unnecessary layer of planning, not required during objective data-oriented model development.
• AI models can pose challenges in identifying and addressing emergent ethical concerns, especially when product teams have not received effective training on common ethical implications that many models face.
• The absence of authoritative ethical standards for AI and technology use more broadly within companies poses challenges for product teams in determining what practices are considered ethical and responsible. Conversely, this can also be a sign that your organization lacks the diversity of thought or experience to consider ethical policies and safeguards.

The incorporation of ethical AI practices is crucial for responsible and trustworthy AI development. For many of the challenges, AI governance software advancements allow companies to govern, monitor and audit models continuously, providing right-time evidence and documentation that demonstrates AI safety and compliance to various stakeholders.

See also: Beware the Dark Side of AI

Companies That Prioritize Ethical AI Principles

Your AI ethics should be aligned with your corporate ethics, standards and practices. If you have ESG policies, seek alignment between those and your AI. Do not view AI in isolation from broader societal values your organization has or is developing. 

Regulated industries such as banking and insurance are familiar with assessing the performance, robustness and compliance of their algorithms and models against standards and controls. They have been doing it for decades. Rapid innovation and AI have forced these industries to streamline and automate these processes to explain their AI continuously for compliance with industry standards.

Some AI-led insurtechs are going as far as to publicly share their audit process and timing. This is a practice that will become increasingly important to discerning vendors, partners and customers who choose third parties to incorporate human-like AI experiences in their products and want to do it ethically and responsibly.

Customers Decide on Ethics and Trust

Your company and your customers have core business ethics to adhere to and uphold. With proper consideration, your ethics for developing and implementing AI will follow.

By building ethical AI principles into your core product strategy, your company can build immediate trust with end users and customers. Leading ethically with AI also ensures that you are building products that don’t become distrusted or misused or, worse, unsafe tools on a customer’s shelf.