Tag Archives: ai

3 Ways AI Can Boost Customer Retention

The insurance industry has steadily been digitizing in recent years. It is taking advantage of technological developments in automation, offering apps to clients and introducing things like electronic proof of auto insurance.

PwC’s Annual Global CEO Survey in 2020 identified customer experience (CX) and core tech transformation as the top two opportunities for companies to set themselves apart, with CX significantly ahead. When customers are communicating with their insurance company, they are usually dealing with an awful experience (a car accident, a medical issue, a home invasion, a roof leak or a full-on natural disaster). Hitting any snags when it comes to getting the service they need is a big factor in driving them to a competitor.

Artificial intelligence (AI) addresses both customer experience and tech transformation and has helped insurance companies improve their services and stay competitive in a tough industry. AI enables many services and processes to be automated, resulting in both cost and time savings. But AI deployment also benefits customer experience, in three specific areas.

Improvement of policies, products and processes

The retail sector has capitalized on custom products and experiences, and the concept that one or two sizes will fit everyone is fading quickly in other sectors, too. The insurance industry is now also latching onto customization, as AI enables companies to leverage data to personalize a core product for an individual customer. 

A company’s AI model can use a client’s historical data to calculate with a high degree of probability that a particular product or policy will be the best fit. In the eye of the customer, they get an attractive product without needing to spend a lot of time on consultations with the broker. For the insurance company, the efficiency gains are impressive, and agents don’t need to spend a lot of time finding the right product for the client.

See also: How AI Powers Customer Contacts

Not only does AI improve the speed of crafting policies for customers, the technology can also speed up underwriting, as well as the claims process — two key touchpoints where turnaround time is essential for CX. If a health insurance customer is filing a claim for an expensive prescription, they will want a simple and quick resolution. Long waiting times may drive them to a competitor known for quicker reimbursements. But AI can bring a competitive edge to an insurance company if customers know they can expect trouble-free claims processing.

Enabling rapid online assistance

The days of customers playing phone tag with agents to get the information they want are long gone. Insurance companies can leverage AI on their platforms — both web and mobile — to allow customers to quickly find an answer to a question. The ability to more easily respond to inquiries from policyholders is especially important following a major event, such as a tornado or hurricane, when there is sure to be a high-volume of online interactions.

AI-powered chatbots have become the first touchpoint for customers in many sectors. When done right, they can give a major boost to an insurance company’s customer experience. Organizations can also save money by automating simple and routine online customer interactions, leading to another win-win situation where customers can quickly get to the information they need.

See also: Wake-Up Call on Ransomware

Using AI to test AI models

Once an insurance provider has an AI model in place to help with the crafting of policies and settling of claims — and the organization begins using it as an integral part of its daily routine — it is vital that the AI model suitably addresses four risk factors: accuracy, stability, flexibility and ethics. AI can be used to test these models against those factors.

Any AI model an insurance company employs should have a high accuracy score. Smart AI model testing will ensure that the results are reliable — with an agreed tolerance for the model, to protect a company’s profitability while appropriately managing their risk. Because data changes over time, testing will ensure that an AI model remains stable when there’s a change in the data in the ecosystem, or in the way the insurance company handles business. Such testing will also ascertain that the model is flexible enough to react to those changes while remaining accurate. 

Above all, AI can be used to test whether the insurance company’s AI models are ethical — meaning that they are not biased toward or against any specific groups of society. Even the largest dataset imaginable can be flawed or biased depending on the data that is included. Therefore, it is vital that these models be tested on a regular basis, to verify that the risk factors are being appropriately applied without bias, ensuring the company’s reputation and its brand.

Giving customers a reason to stay

As the saying goes, it’s cheaper to retain a current customer than to attract a new customer, so insurance companies need to look at technological innovations that can improve customer retention. By recognizing the value AI can bring to all aspects of the customer experience, insurance providers can deliver fast, accurate and fair service to policyholders. 

Whether a business provides health, car, home or other types of insurance, there will always be a line of competitors just waiting to snatch away disappointed and frustrated customers. What’s most important is for insurance companies to focus on giving customers reasons to stay.

Top Problems That AI, ML Help Solve

The global life insurance and retirement industry is facing an inflection point due to the convergence of challenging economic, technological, competitive and societal headwinds. Product-driven business models of the past will not be sustainable because insurers cannot adapt quickly enough to changing customer needs. This problem is on top of mature markets, strict regulatory requirements, low interest rates and tight margins. The COVID-19 pandemic has made it even more urgent for life insurers to redefine their role, take bold measures and address these changes.

The good news is that many global insurance leaders are already making large investments in digitization, innovation and cultural change. Going digital has been a top priority, as it helps reduce cost and enhances customer experiences, leading to the increasing adoption of predictive analytics, artificial intelligence (AI) and automation in various business functions in the industry. According to McKinsey estimates, the potential total value of AI and analytics across the insurance vertical is approximately $1.1 trillion.

Soon, AI will be deeply embedded into the insurance value chain, providing unmatched power to insurers: automating manual processes in underwriting, eliminating errors and inefficiencies in claims processing and enabling predictive insights to deliver superior outcomes. Below are the top challenges that AI and machine learning (ML) will help solve in the insurance industry. 

1. Underwriting and Pricing — While pricing personal auto policies is mostly automated today, the underwriting process is still manual for commercial property. For commercial property insurance, the underwriter needs a lot of information, such as occupancy, data on adjacent buildings, loss estimates and typical hazards. Some of the data may be available online but may be outdated and might require onsite verification. This is why human judgment is critical. A PwC report on top insurance issues noted that carriers are devoting considerable attention to helping underwriters use models and AI-driven tools to supplement their knowledge. Underwriters are becoming increasingly comfortable marrying what they’ve learned from personal experience with insight from models to make the most informed decisions possible. Soon, underwriting will be fully automated, supported by machine learning models that ingest vast amounts of data through an ecosystem of vendors. 

2. Claims Processing — In the future, machine learning algorithms will manage claims routing, increasing efficiency and accuracy dramatically. According to a McKinsey report, claims for personal lines and small-business insurance will be fully automated, enabling carriers to achieve straight-through-processing rates of more than 90% and dramatically reducing processing times from days to hours or minutes. Unlike with the traditional practice, involving manual methods of first notice of loss, the burden will no longer be on the customer to inform the insurance carrier about an event. The process will now be automated, relying on  IoT sensors and real-time monitoring to prevent incidents from happening and sending notifications for critical events requiring immediate attention. An app on a smartphone will handle all interactions, with the capability to trigger claims automatically upon loss. Other technologies will support claims processing, such as natural language processing, deep learning and text analytics. 

See also: Wake-Up Call on Ransomware

3. Fraud Detection — Insurance fraud can cost companies millions to billions of dollars, as there are thousands of claims filed every day. Assigning insurance agents to investigate each case will be time-consuming and expensive. Using AI, insurers can evaluate millions of documents and data points in record time. They can cross-reference several databases and incorporate multiple external data sources, which would be impossible without automation. Anomaly detection models can identify deviations and flag cases for review. Leveraging learnings from previous fraud cases and using real-time data, AI and ML models can identify threat signals before they might become a more substantial problem. 

4. Other Use Cases — A common use case is using predictive analytics for estimating policy cancellations. Customer churn is one of the most problematic aspects of customer management for insurance companies. When high-value customers churn, insurance companies often replace existing businesses with new, more costly customers that lower profitability. Creating AI and ML models that can accurately forecast churn behavior can boost profitability and revenues. 

As insurance carriers get better at leveraging data and implementing predictive analytics, the focus will shift from product-led to customer-centric models. The insurance industry’s adoption and investment in digital capabilities to unify data, advanced analytics and people will ultimately make the industry more agile, efficient and transparent. The winners that go above and beyond will start to offer personalized products based on individual customers’ unique needs and enhanced customer experience.

Future of AI and ID Management

Identity management has been an obstacle for commercial insurance companies for a very long time. Many thought that problems would dissipate or at least become easier to correct by moving to digital systems, but, in reality, identity management has only grown more complex. It is obvious that we need a better way.

Now, there is fresh hope that identity management will become much easier to wrangle. Artificial intelligence (AI) is progressing rapidly, to the point where it could become a tremendous tool in identifying and cleaning up inaccurate data as well as linking the right providers to the correct claims.

Let’s take a step back and examine the key issues in identity management today to understand how AI could be used to shore up existing gaps and move the industry forward.

The Data Problem

First of all, by identity management, as it is applied to insurance, I am referring to a special case of entity resolution, i.e., the process of linking references to providers in claims, bills and other data to a single flesh-and-blood provider — the so called “single belly button.” This is facilitated by maintaining a dataset of the actual providers working all over the country, with their names, addresses, specialties and networks — essentially all the data associated with them for billing purposes. These “golden sets” also are available for attorneys in the claims space, functioning nearly the same way. Still, for the sake of clarity, I’ll focus on medical providers in this article. These lists are available through a handful of third-party vendors (and certainly some organizations have developed their own), and they must be constantly updated as the ground truth evolves.

See also: Intersection of AI and Cyber Insurance

The Missing Link

Currently, numerous different golden sets have varying degrees of accuracy and cleanliness. While this is certainly problematic, the real challenge in identity management is the linkage process itself. This is because much of the provider references in the claims, bills and other data can be considered stale or dirty.

There are myriad reasons for stale and dirty data. Doctors change their name through marriage or for other reasons; they move to other cities; they might add a specialty or change focus, Joe Smith might become Josephine Smith. All of these things and more make the process of linking these references to the correct provider very difficult. In many cases, today’s systems lack the ability to link references in claims to golden sets; instead, linking falls to claims representatives. One of the biggest identity management tasks remaining today is the ability to uniquely and accurately link a claim to the right provider with the correct billing information.

Many companies try to build their own link, but it has not been smooth sailing. Developing such functionality is an expensive, time-consuming, complex endeavor. Without clean, accurate, linked datasets, claims can go wildly off track. But there is hope.

AI Will Fill the Gap

AI has shown its effectiveness in improving claims operations processes, pulling out key insights to resolve claims quickly without attorney involvement. Now AI could be applied to solve the linkage problem as well.

AI systems that aggregate data from actual anonymized claims, bills and other data throughout the industry could be used to read massive volumes of data, recognize pattern and find the links between specific providers and claims. Systems could be trained to identify and update records, managing identities persistently and in real time.

Imagine just for a moment that you had a very high threshold of confidence in identifying the correct provider for a claim and that the provider automatically would be issued a unique ID (in the U.S., that of course is the National Provider Identifier, or NPI) that stays with him or her throughout the life of the claim so that every time a change is made — a note filed, a bill paid — the correct provider at the correct location automatically comes up. No detective work, no guesswork.

This is now possible from a technological standpoint, as we have seen in creating CLARA’s solution. I can attest that it requires a significant investment of time, effort and intellectual property to build in-house. Given the rate of AI advancement, market adoption and pressing industry need, there is no doubt that it won’t be long before nearly all identity management systems are powered by AI and machine learning technologies.

See also: Insurance Outlook for 2021

As I hope I have shown, the data available to the industry today is nowhere near sufficient. The bar for identity management — and therefore the level of investment, skill and innovation applied to this problem — will continue to increase. Those organizations that prepare to embrace new applications of AI for identity management will be the ones that thrive and modernize claims, driving down costs and increasing efficiency. The companies that resist this transformation will get left behind as they struggle to sift through their dirty, messy data.

As first published in Digital Insurance.

Key to Transformation for Auto Claims

The word “transformation” is overused, and yet here in the auto insurance claims industry there is no better word for a process that is being changed so dramatically from beginning to end, and at every step in between. 

But real transformation, while claimed by many, is in reality only enabled by the exceptional few. That is because transformation occurs through a collective, inclusive effort, not a silver bullet technology. And complete transformation requires the active participation of the end-user, to ensure higher levels of acceptance and satisfaction. Transformation must be good for the business and the customer, or it will likely not take hold at all. 

Foundational to Success

Digital transformation is the essential driver behind how companies will add value and deliver services to their modern customer, a customer who expects and trusts digital interactions. AI is critical to processing and assessing all inputs and removing friction. Yet AI alone cannot deliver transformation. 

Let me explain.

We know data availability is increasing rapidly across multiple dimensions – volume, velocity and variety. In the last two years, more data was created than in the entirety of human history. This is a fraction of the data that will be available in the near future as connections continue to multiply, becoming increasingly bi-directional and informing virtually everything. Right now, there are more than 50 billion connected devices in the world, and connected cars are emerging as an important digital platform.

Artificial intelligence is the only way businesses can leverage the tremendous amounts of data available. AI synthesizes all of this data. This is positive and necessary. But AI output is often delivered to humans, reviewed offline and paused before actions occur. Companies have to eliminate this pause and disconnect in the process to transform their operations. AI decision-making must be digitally connected to operating systems or consumer interfaces or both to drive action and to create a truly elevated, digital experience.  

See also: Transforming Auto Claims Appraisals

Relevant mobile technologies, network connection management and industry-specific workflow applications are required to activate AI, automating tasks based on that data to speed up and simplify lengthy and complex processes. The auto insurance claims process is an ideal candidate for such transformation. Our industry needs to connect AI to technologies that drive action. 

Here’s an example of how a transformed auto claim experience can look to your policyholder when AI gets put into action with mobile and network technologies:

  • Pat enters his vehicle in the morning. The app on his phone activates and begins tracking his trip so that his auto insurance policy premium is calculated for only the time he is in transit, based on the policy he selected upon enrolling. 
  • On arrival at his employer’s office parking lot, Pat accidentally scrapes the side of his vehicle on a pillar. 
  • His vehicle and app automatically detect the incident and offer Pat the opportunity to submit the incident to his insurer to determine if a physical damage claim should be opened.
  • Pat decides to proceed and immediately receives a text link with instructions about how to take a few smartphone images of the damaged area and text them to his carrier. 
  • Pat is immediately notified by text that the damage is minor and that the car can be safely driven but that the repair cost likely exceeds his policy deductible by at least $500. 
  • Pat decides to file the claim and receives a text with a list of nearby repair facilities, including consumer ratings, shop certifications or specialties and availability. 
  • He taps a few links and schedules the repair, and once he arrives a pre-arranged temporary rental car will be waiting for him. 
  • Pat continues to receive status updates from his insurer until he is advised what time his vehicle will be ready for pickup or delivery, as preferred.  

Note that the steps described above begin and continue with AI-enabled decision-making and workflow management. Out of view of the policyholder, AI and digital connections are powering the parts ordering process, and the repair facility is digitally paid by the insurer within hours of the vehicle being delivered. Without these enabled technologies, a digital end-to-end experience would not be possible. But when combined with the other elements, the result is transformative, completely digital.  

Sourcing the Data That Powers AI and Drives Decisions

The relationship between the ability to reliably predict outcomes and the absolute volume of historical claims data leveraged to train the software is directly proportionate – the greater the amount of relevant data used, the better the outcome. We frequently hear from our insurance clients of all sizes that the volume of data needed to develop reliable algorithms is greater than even the largest insurers have available. CCC has processed more than $1 trillion of claims-related data, which we put to work to develop hundreds of actionable AI models. And while data relevancy is essential, another key difference in AI efficacy is the use of a combination of AI disciplines. Deep and machine learning and business rules combine to deliver the most reliably predictive, comprehensive results for faster, smarter resolutions. 

Here’s how: 

Deep learning is an AI method that uses historical data to inform which action is likely to lead to which outcome. Let’s take photo-estimating as an example. To train an AI model that can review smartphone images from a collision and predict whether a vehicle is repairable versus a total loss, the AI model needs to learn from historical data: photos of other car crashes, as well as the claims data that accompanies those photos regarding the parts, labor, cycle time and medical assistance needed for each claim. The question is: Does the AI model have enough historical data to make that prediction actionable? A few hundred images are helpful, but decades’ worth of wrecked car images and related metrics make the AI model far smarter. Another question: Can the AI model sort out the anomalies from the requisite data set? Can it learn from them?

Another key discipline is machine learning, which allows historical data to be influenced by behavioral or pattern changes that might make recent actions more likely to occur again. Let’s say you have been a Facebook visitor every day for the last five years, but more recently you’re only visiting Instagram. In this case, the majority of data would say you’re going to visit Facebook again, but recent activity would suggest Instagram is a better prediction. Why does this matter? Vehicles and parts are not static. New cars and parts are introduced continuously; if an AI solution is going to be effective, it needs to base predictions on data that can account for recent behaviors, not just historical data. 

A less sophisticated, yet foundation disciple, includes the use of rules. A rules-based approach can offer helpful predictions when historical data is not available or recent activity is not accurate enough to ensure a reliable prediction. Suppose that an inbound technical support email contains the word “urgent” in the body. A rule is triggered, and that email is forwarded to someone who can immediately act on it. These types of rules can get into extremely complex decision points, leading to hundreds of potential rules, some of which may even conflict with each other. This is why rules-based AI is an incomplete approach that can fall short in accuracy and reliability. Yet, because data and domain experience aren’t required to create rules-based AI, it is a helpful starting point that can assist companies to begin the journey of automating complex workflows such as auto insurance claims.  

See also: Auto Claims: Future May Belong to Bots

When It All Comes Together – A Reimagined Insurance Experience

When the claims experience is working in harmony as a result of automated, AI-enabled decisions and all the needed inter-company, inter-industry integrations, not only will the insurer’s customer’s experience be maximized but real hyper-personalization can be achieved, meaning that each insurer’s individual customer will enjoy an exemplary service experience in the manner and method that they expect and prefer. 

Industry transforming technology is here and ready to be combined in time to meet consumers’ evolving expectations. Insurers are in a position to connect AI, mobile and network to transform what’s possible.

The Promise of Predictive Models

An innovation strategy around big data and artificial intelligence will uncover insights that allow smart carriers to acquire the most profitable clients and avoid the worst. Companies that develop the best portfolios of risks will ultimately enjoy a flight to quality while those left behind will compete for the scraps of insurability.

Insurers are also trying to individualize underwriting rather than use the traditional underwriting of risk categories.

As such, the insurance industry finds itself in a data arms race. Insurance carriers are leveraging their datasets and engaging with insurtechs that can help.

For the underwriter, big data analytics promise the ability to make better decisions with respect to risk selection and pricing. Underwriters have thought too many times that if they had just understood a particular area of risk better they would have charged a lower price and won the business; or had they had that little extra piece of information they would not have written an account that turned out to be unprofitable. Most certainly, underwriters would assert that with better information they would have charged a more appropriate price for a risk and most definitely would not have lost money.

One solution has been developing predictive underwriting risk selection and pricing models. By leveraging datasets previously unavailable, or in formats too unstructured to use, algorithmic models can better categorize and rank risks, allowing an underwriter to select and assign the most appropriate price that rewards better risks and surcharges those that are riskier. Better risks might be those that are simply less inherently risky than others (e.g., a widget manufacturer vs. an explosives manufacturer with respect to product liability or property coverage), or those whose behaviors and actions are more cautious. Through a predictive, data-driven
model, underwriters will be able to build profitable and sustainable portfolios of risks, allowing them to expand their writings to a broader customer base, pass along cost savings from automation to their clients, provide insights into means by which their insureds can reduce risk or identify new areas of coverage and product and bring more value to customers.

With this win-win situation at hand, the insurance industry has charged forward in data mining the decade’s worth of their own internal information, as well as accessing public databases, leveraging data brokers and partnering with insurtechs that have their own data lakes they can access. Algorithmic models then are being fine-tuned by actuaries, statisticians and behaviorists to find causation links and correlations between seemingly disparate data points with the intention of divining future loss outcomes. In this digital frenzy, what gets lost, however, is that there can be social costs from the methods by which all this data is used.

See also: 11 Keys to Predictive Analytics in 2021

Balancing Social Good With Social Cost

It is not false altruism to reward good risks, build resiliency in portfolios or discover insights that lead to new products and services. However, underwriters must recognize that they are inherently in the business of bias. While it is acceptable to be discerning between a safe driver and reckless one, it is unacceptable to build into underwriting decision a bias toward race and religion and many times gender or health conditions. It is therefore essential that underwriters, and the actuaries and data scientists who support them, act responsibly and be accountable for any social failures of the algorithmic models they employ.

With our predictive risk selection model in mind, consider some of the available data that could be processed:

–decades of workers’ compensation claims data

–policyholder names, addresses and other personally identifiable information (PII)

–DMV records

–Credit scores and reports

–Social media posts


–Wearable tech data

–Biometric data

–Genetic and genealogy information

–Credit card and purchasing history

Consult algorithmic accountability experts like law professor Frank Pasquale, and they will provide you with additional data sets you might not even know existed. Professor Pasquale described the availability of databases of everything from the seemingly innocuous (wine enthusiasts) to those that shock the conscience (victims of rape). With the myriad of data available and so much of it highly personal in nature, underwriters must recognize they have a responsibility to a new set of stakeholders beyond their company, clients, shareholders and regulators — namely, digital identities.

The next risk of social harm is in how that data is used. Predictive models seek to identify correlations between new points of data to predict loss potential. If correlations are wrong, not only could it jeopardize the underwriter’s ability to properly price a risk, but the correlations could result in an illegal practice like red-lining. This situation could occur accidentally, but a dataset could be used nefariously to circumvent a statute prohibiting use of certain information in decision making.

In California, there is a prohibition on using credit scores in underwriting certain risks. Perhaps a modeler for a personal lines insurance product draws information from a database of locations of check cashing stores or pawn shops and codes into the algorithm that anyone with an address in the same ZIP code is assumed to have bad credit. You would hope this would not happen, but insurance companies use outsourced talent, over which they have less control. Maybe a modeler works outside the U.S. and is innocently unfamiliar with our social norms as well as our regulatory statutes.

There are also social risks related to speed and complexity of predictive models. Dozens of datasets might be accessed, with different coded correlations and computations processed that are then weighted and ranked until a final series of recommendations or decisions are presented to the user. Transparency is difficult to attain.

If there is something ethically or statutorily wrong with a model, the speed at which processing can occur and the opaqueness of the algorithms can prolong any social harm.

Don’t Throw the Baby Out With the Bathwater

While regulation of big data analytics is not well-established, there are governance steps that insurance companies can take. Insurance companies can start by aligning their predictive models with their corporate values. Senior leadership should insist that decision-making technology adhere to all laws and regulations, but more generally will be fair. Fairness should apply to the process and to the rendered decisions. Standards should be established, customers treated with respect, professional obligations fulfilled and products represented accurately.

Insurance companies should audit their models and data to ensure a causation linkage to underwriting loss. Any data that does not support causation should be removed. Parallel processes employing traditional and artificial intelligence techniques should also be run to confirm that an appropriate confidence level of actuarial equivalence is met. Data should be scrubbed to anonymize personally identifiable information (PII) as much as necessary to support privacy expectations and statutes. To remove biases, audits should identify and require exclusion of information that acts as a proxy for statutorily disallowed data.

In essence, the models should be run through a filter of protected class categories to eliminate any illegal red-lining. Because models are developed by humans, who are inherently flawed, modelers should attempt to program their machine learning innovations to identify biases within code and self-correct for them.

From a base of fairness, carriers can take steps to promote transparency. By starting with an explanation of the model’s purpose, insurers can move toward outlining the decision-making logic, followed by subjecting the model to independent certification and finally by making the findings of the outside auditor available for review.

Insurers can look to trade associations and regulatory bodies for governance best practices, such as those the National Association of Insurance Commissioners (NAIC) announced in August 2020. The five tenets of the AI guidelines promote ethics, accountability, compliance, transparency and traceability.

See also: Our Big Problem With ‘Noise’

One regulation that could be developed would be imposing rate bands. Predictive engines would still reward superior risks and surcharge poorer-performing accounts, but rate bands would temper the extremes. This regulation would provide a balance between the necessity for mutualization of risk and individualization of pricing that could lead to unaffordability in certain cases.

Finally, insurance companies should recognize the importance of engaging with regulators early in the development of their AI strategies. A patchwork of regulation exists today, and insurance companies could find regulatory gaps that they might be tempted to exploit, but the law will catch up with the technology, and carriers should build trust with regulators from the onset, not after a market conduct exam identifies issues. Regulators do not wish to stifle innovation, but they do strive to protect consumers.

Once regulators are comfortable that models and rating plans will not unfairly discriminate nor jeopardize the solvency of the carrier, they can help enable technology advancements, especially if AI initiatives facilitate an expansion of the market through more capacity or new products, lowers overall market costs or provides insights that helps customers improve their risk profile.

In the data arms race that carriers are engaged in with each other, better risk selection and more accurate pricing are without question competitive advantages. Another, often-overlooked competitive advantage is an effective risk management program. Robust management of a company’s AI risks will reduce volatility in a portfolio and promote resiliency. With this foundation, a carrier can deftly outmaneuver competition and should be an additional strategy that is prioritized.