AI Pricing Models in Insurance: Proven Tech, Measurable Gains

Delivering Growth, Risk Mitigation, and Enhanced Profitability

We are awash in stories about artificial intelligence (AI) and its applications in all corners of our lives, both personal and professional. The accounts run the gamut from ecstatic to apocalyptic and everything in between.

Google the term “AI” and you’ll get over 1.6 billion results. Narrowing it to “AI in business” reduces the result set to 1.2 billion. Narrowing it to insurance still yields 193 million references. The result is 79.5 million references when one narrows the search further to “best practices for AI in insurance.”

Clearly, there is no shortage of information on the topic of AI in insurance.

What decision-makers need is a straightforward discussion of where AI can be applied to insurance pricing models, its key benefits, and quantifiable outcomes. This blog post sets out to do just that.

An Overview of AI Pricing in Insurance

Insurers are no strangers to discussions of the application of AI to their businesses. As far back as the 1980’s they began to explore the use of machine learning (ML), a subset of artificial intelligence, for demand modeling. Since then, AI in its various forms has found its way into the mainstream of nearly every aspect of the insurance sector, including risk assessment, pricing, rating, and underwriting.

The possibilities are immense. McKinsey, for example, estimates that the application of artificial intelligence will drive up to $1.1 trillion (yes, with a “t”) in value in the insurance space in the foreseeable future.

While early adopters gravitated to AI due to the lure of new technology, the vast majority of carriers are in search of solid business reasons for its application. AI pricing models top that list.

Insurers are looking to take advantage of the power of Intelligent Operations, or Intelligent InsurOps, which will allow them to tailor every customer interaction to the individual, driving faster revenue realization, higher customer satisfaction, and increased profitability.

The Insurer’s Pricing Dilemma

Attempting to optimize pricing in insurance is as old as the profession itself. Finding the optimum balance of risk and reward has bedeviled insurers for centuries. Using computing resources for the exercise dates to the earliest use of computers themselves.

To this day, insurers wrestle with several key issues in honing their pricing models:

1. The Complexity of Existing Models

Built up over decades, the pricing models of many insurers are a mish-mash of “spaghetti code” that may or may not be well-documented and/or understood by the teams that work with them on a daily basis.

Those models require intricate coordination across business functions to work best, yet the responsibility for devising, coding, testing, and deploying them into production is often split across pricing, underwriting, analytics, and IT, introducing too many opportunities for mistakes, miscommunications, and re-work.

Add this all up and processes grind slowly, and the handoffs expose the carrier to unnecessary and dangerous financial and regulatory risks.

2. Drowning in a Sea of Data

By some accounts, data is the most valuable commodity on earth, more valuable to modern economies than oil, gold, or diamonds. Like those previous standards of wealth, data is useless unless properly extracted, refined, and shaped for high-value purposes.

The problem with data for insurers is not one of scarcity, but instead one of overabundance. Internal data on current customers, claims, lost business, pricing sensitivities, and written policies alone is overwhelming.

Add in external data such as economic projections, demographics, climate data, and competition, and separating the wheat from the chaff becomes nearly impossible without the application of modern analytics.

3. Lack of Personalization

Combine complexity and an inability to effectively use data to advantage, and the result is a lack of personalization in the offers that insurers can make. Personalization expectations go well beyond that which demographics alone can supply. True personalization considers more than age, income, geography, etc.

Accustomed to personalized offers in every other aspect of their economic lives, consumers expect nothing less from their insurers. They have come to expect personalization at every step of the customer experience (CX): marketing, initiation, claims management, policy renewal, and offers of new or complementary products and services.

Insurers who have not transformed their pricing processes simply cannot offer customized products and services, pricing, and discounts that make customers and prospects feel like they are being addressed as individuals.

4. Reaction vs. Anticipation

The above effects snowball into inaction and reaction to the market, rather than anticipation of what could happen next.

Complexity slows the insurer down, making it less agile and nimble, and opening it up to attack by competitors who have mastered modern methods and technologies.

Many insurers are mired in processes that require months to implement new pricing strategies, by which time their competitors have revised and updated their pricing many times over, and used that advantage to capture more business.

The ability to respond in real-time to market changes and emerging trends is the hallmark of the modern insurer. Their lack makes carriers “sitting ducks” for attack by the competition.

5. Resource Intensiveness Eats Away at Profitability

Add up all these challenges, and insurers who are “behind the curve” in transforming their businesses spend too much time, effort, and resources to even “run in place.”

Their people are stressed, inefficient, and all too aware that they and their firms are falling further behind by the day.

On the IT front, many insurers’ pricing efforts still remain woefully dependent on legacy systems first implemented decades ago. Those systems require inordinate amounts of time and energy to maintain, and will never serve as a source of competitive advantage.

Stakeholders, including shareholders, see the slide and react accordingly, driving down the market values of laggard insurers.

Corporate leadership looks at all these barriers, and their overwhelmingly negative effects, and desperately searches for a way out.

AI Pricing Models – How We Got Here

Humans have been exhibiting analytic skills for millennia. They have progressed, well before computing resources became available, from the descriptive (“the wolf attacked the tribe”) to the diagnostic (“Was it because the wolf’s cubs were nearby?”) to the predictive (“when will the wolf attack again?”) and finally to the prescriptive (“what can we do to make sure the wolf doesn’t attack again?”).

Insurers initially began employing computers to automate repetitive and administrative tasks (billing, claims processing, accounting, financial reporting, etc.) and were among the pioneers in applying computing to business operations. Many of those early administrative systems are still in place today and functioning quite nicely for those particular workloads.

Insurers soon discovered that there was additional value-add in applying those computing resources to tasks such as risk prediction, underwriting, claims analysis, and customer management, resulting in the first uses of predictive analytics in insurance.

While a step forward in the ability to process and utilize large volumes of data, these early solutions relied on what now are considered “brute force” methods, such as regression analysis and decision trees. Over time these models increased in complexity and efficacy, but also became more unwieldy and difficult to maintain and update on a regular basis.

AI Pricing Models Today

The dawning of the 21^st century and the advent of the Internet forced the hands of insurers.

Customers and prospects have come to value speed, personalization, and flexibility in their dealings with carriers. No aspect of this relationship is more important than pricing, and the presentation of policy options tailored to consumers.

The competitive landscape is unrelenting. Non-traditional vendors have also entered the fray, in many cases unencumbered by legacy technology and with the ability to begin with a “clean slate” in how they present policy options, close business, and develop and promote their offerings.

The Earnix Solution

The Earnix solution capitalizes on all these advanced technologies to bring the best in AI pricing to insurers.

In today’s world of hypercompetition, AI insurance pricing enables the rewarding experiences that customers and prospects crave. Pricing teams can deploy dynamic pricing that moves at the speed of the market, and pricing models can be scaled up and maintained with ease.

These agile insurance solutions integrate with existing systems and infuse automation and industry-leading analytics into every aspect of the pricing process.

The heart of the solution is Earnix Price-It^TM, a single-platform solution that includes integrated modeling, simulation, fine-tuning, and deployment, allowing insurers to bring personalized and innovative products to market faster while maximizing ROI on their technology investments and ensuring strong governance and regulatory compliance.

The Benefits of AI Pricing in Insurance – Delivering Real-World Results

The Earnix solution is the answer to the challenges we outlined above.

Let’s examine the benefits, and delve into some examples of real-world Earnix results.

1. Enhanced Accuracy and Rock-Solid Risk Assessment

Insurers’ reputations and businesses rest on their well-deserved track records of meeting the stringent risk and regulatory requirements their constituencies demand. With Earnix, they can leverage vast amounts of internal and external data, make more accurate predictions, and drive additional revenue, all without taking on added risk.

Earnix allows carriers to tap into the intelligence needed to align pricing and product strategies, uncovering hidden patterns of customer and market behavior critical to growth. Insurers can align to real customer risk and accelerate organizational growth, offering up pricing that attracts business while remaining strong on the risk mitigation front.

Insurers can achieve that elusive balance, and use pricing as a competitive weapon, as outlined by McKinsey: “Price remains central in consumer decision making, but carriers innovate to diminish competition purely on price.”

Insurers who embrace AI can avoid a “race to the bottom” that would be driven by their most efficient and cutthroat competitors.

A case in point – a leading US auto insurer, an Earnix customer, has shown a 15% increase in quote conversions, without taking on additional risk. The best of both worlds.

For another perspective, check out this customer case study, in which Earnix is described this way, per the insurer’s Chief Actuary:

"We immediately saw that the Earnix solution is a modern, comprehensive solution and would give us all the tools we needed to manage our entire pricing approach. The Earnix Price-It platform is analytically sophisticated, completely connected, and easy to use."

Or, as Michael Chang, the Chief Risk and Underwriting Officer at Hollard put it:

“The Earnix Rating Engine stood out with its ability to robustly integrate rating and pricing processes across Hollard. With the Earnix platform, we expect to benefit from improved pricing controls, with real-time model deployment agility, in-platform modeling capabilities, and advanced scenario planning features.”

2. Vastly Improved Product and Pricing Personalization

Insurers need not cede the personalization high ground to digital upstarts or mourn the head start that some of their more aggressive traditional competitors may have built in this area.

Every insurer can perform like digital-native industry leaders. Being able to define, personalize, and deploy AI-powered offers and bundles across their enterprises creates the ability to provide a more streamlined customer journey that in turn improves the bottom line.

The ability to deploy new insurance product personalization configurations faster can lead to an increase in cross-selling, upselling, and total sales. Carriers can consistently deliver a range of the most appropriate insurance packages to customers.

As McKinsey outlines it, when personalization is done well, from consumers’ standpoint, “Pricing is available in real-time based on usage and a dynamic, data-rich assessment of risk, empowering consumers to make decisions about how their actions influence coverage, insurability, and pricing.”

Earnix technology has enabled leading insurers to use personalization as a competitive weapon, as outlined by David Cummings, Senior Vice President, Chief Actuary & Head of Analytics at USAA.

As he puts it in this presentation, “We rely on data and analytics and our culture of customer centricity to guide us through change at USAA.”

3. Enhanced Real-Time Adjustments and Flexibility

Gone are the days when times-to-quote measured in days were acceptable in the insurance industry.

Customers are increasingly dealing directly with insurers on their websites and expecting instant quotes. Producers, likewise, will no longer wait for insurers to provide quotes in anything slower than real-time. Both groups are more than willing to move on if their expectations are not met, and met consistently.

Changing market conditions or competitive actions can no longer be tolerated. Too much data, or data that cannot be analyzed quickly and thoroughly enough, can’t stand in the way, either.

Insurers need the ability to innovate in a mode that gives them maximum flexibility, to bring quickly and efficiently to market the most accurate and productive analytical models – models that leverage the full power of their teams’ analytical skills.

The results can be game-changing. For example, one Canadian insurer cut its rate-to-market timeline from six weeks to four days.

A US carrier, focused on the small business market, has adopted Earnix and shifted its modeling from general linear models (GLMs) to ML in only three months, and is now able to update, test, and deploy new pricing rules in less than an hour.

Only AI pricing and quoting can deliver those kinds of results.

4. Cost Efficiency and Resource Optimization

Applying AI to insurance pricing provides a powerful source of cost efficiency and resource optimization. The key is to find a way to implement new technology that allows for instantaneous adjustment to changing market conditions, while not “throwing the baby out with the bath water,” leveraging in-place systems that still serve a useful purpose.

The Earnix solution is designed to work in tandem with still-useful systems and layer on new capabilities, resulting in a complete - and completely flexible - IT infrastructure.

Earnix’s cloud-based solutions dramatically reduce the cost of achieving the goals we’ve outlined, while also reducing the transition risk and time required to capitalize on AI pricing models.

Earnix helps ease the transition. Just one example is the insurer BavariaDirekt.

With Earnix, they have been able to

• Develop and deploy new online pricing offers much faster

• Overcome reliance on third-party systems, and

• Simplify key pricing simulations and calculations

Because Earnix is an open analytics platform, teams can capitalize on all aspects of their skill sets and can use a variety of third-party tools. One such complementary tool is Guidewire.

When the two toolsets are used in tandem, productivity can skyrocket. For example, across a range of insurers these results have been reported:

• One insurer has consolidated 8 separate rating engines into Earnix

• Another has shifted 170 people from testing to higher-value work

• And an additional carrier has experienced 10x faster quoting than the incumbent solution the Earnix/Guidewire solution replaced

Conclusion

We’ve seen how AI pricing can propel carriers into a new world, a world of Intelligent Operations, in which they can compete more effectively for new business and maximize the retention of a loyal customer base. This in turn increases revenue, reduces cost and overhead, and drives new levels of profitability.

Insurers who capitalize on AI technologies will evolve in a way that McKinsey describes this way: “Insurance will shift from its current state of ‘detect and repair’ to ‘predict and prevent,’ transforming every aspect of the industry in the process.”

Earnix Price-It delivers on the promises of this new world of insurance pricing models.