InsurAnalytics.ai

[insuranalytics-blog]

Virtualizing P&C Claims Adjudication in the New Normal

Challenging Times

We are living in unprecedented times, experiencing a black swan event the likes of which most of us have not seen (and likely will not see in future) in our lifetime. The COVID-19 pandemic is causing unprecedented human suffering and fatalities. Since this is a novel virus there isn’t a therapeutic treatment or vaccine; consequently heeding to calls from local, state and national health care officials, businesses around the world are taking dramatic steps to “flatten the curve” and reduce community transmission in regions hardest hit by the pandemic. Although, this phenomenon is expected to be temporary, the timeframe for recovery seems uncertain at this point; while brilliant scientists around the world work on a vaccine, the availability is potentially 18 months away. Most companies are trying to adapt to the new reality and fine-tuning their business models and processes with the goal to best serve their customers in this challenging environment while ensuring the ongoing safety and health of employees at the same time.

Impact on P&C Industry

The cost of the much needed “social distancing” is fairly dramatic, the economic slowdown is likely to be widespread and deep as a result of direct impact to some industries (small business, travel and leisure) but more importantly second and third order impacts to a number of other industries. As unemployment rises, the P&C insurance industry will likely be impacted in terms of people’s ability to own cars or homes and consequently purchase insurance. We will likely see some decline in earned premiums for most carriers in the short to intermediate term, as acquiring new customers becomes more challenging.

On the flip side, there’s also likely to be an unintended consequence of reduction in churn as more people with insurance will resist change in these uncertain times. Also, with reduced driving across the board, the number of claims is likely to be significantly lower resulting in lower losses for carriers during this phase. Although personal driving is expected to be lower, it’ll likely be balanced by somewhat increased delivery service drivers (many of them individual gig-economy workers with personal insurance). In the bigger picture though, the overall impact on the combined ratio for insurance carriers may be limited to a large extent. Notwithstanding that, insurance carriers carry the huge responsibility of social distancing and ensuring the ongoing safety and health of employees as well as customers; figuring out how to serve customers remotely and virtually is the key to differentiate themselves in this tough environment, digitization efforts by carriers will also accelerate as a result of the current environment. A number of services and functions in a customer lifecycle are already being performed digitally and virtually (with minimal physical contact); marketing, underwriting, quoting, policy and billing are prime examples where minimal physical contact is being practiced already, systems and processes are in-place to run the business smoothly. By far, claims remain to be the most contact intensive part of the insurance customer lifecycle.

Adjudicating Claims in the New Environment

There’s a limited choice for insurers, but to adjudicate claims virtually (contactless, remotely) in the new environment. As you might imagine, there are several challenges in getting all (or the vast majority) of claims adjudicated given that certain activities within a claims lifecycle are contact centric. Activities listed below require adjusters, third-parties, vendors and claimants to meet each other in order to accurately arrive at the truth:

Coalescing inputs from experts and various parties

Systems and processes are not geared for virtual Inspections, investigations and triaging

Communication channels are not mature and geared to enable virtual adjudication causing delays in obtaining and processing information

Managing and validating vendors and their estimates

Inspection of the physical damage

Verifying statements from witnesses and claimants

So, How Do We Get Over These Hurdles?

Most industries and businesses are pivoting and adapting their business processes in order to figure out how to serve their customer’s needs utilizing contactless delivery, virtual mechanisms, etc. Adjudicating claims virtually is not new, insurance carriers have taken steps in the past towards auto-adjudication (low-touch, no-touch) and continue to innovate and move in that direction. The current situation necessitates the acceleration of this phenomenon from a currently small portion (<10% of claims) to significant and substantial part of overall claims, because of COVID-19 and consequently social-distancing. This trend will likely continue well beyond the current situation and become the new norm for leading insurers and way for insurers to differentiate themselves. There are a number of approaches and steps that can help achieve this goal, following are some of the key elements:

As you can see, opportunities exist for carriers to shore up their efforts in adjudicating claims virtually, recent technological innovation has provided a tremendous platform for carriers to leverage and achieve their short term and long term goals.

The Road Ahead

The current COVID-19 situation is one of the most disruptive exogenous event of our lifetime, the human suffering, economic and social disruption are enormous. Every such disaster also brings an opportunity to retool and reinvent to make things better for humanity overall. In order to be prepared for similar or worst situations in the future, carriers need to significantly enhance their current digitization and virtualization of services programs. Accelerating a move to the cloud and AI is the need of the hour and focused spending on technologies that strengthen customer relationships, reduce expenses, and protect employees. For insurance carriers, this is also an opportunity not only to achieve short-term goals like social-distancing by adjudicating more claims virtually, but also achieve longer term business goals of bringing efficiency, speed and expense reduction in their claims operations.

Learn how InsurAnalytics' AI platform help P&C insurers assess overall auto damage cost and predict payout estimation https://bit.ly/3cgILDY.

[insuranalytics-blog]

Data Quality in Claims Adjudication

Claims are one of the most important customer touchpoints for any Insurance carrier and the real test of a carrier’s commitment to quality of service. However, it is also the most labor-intensive operation in the insurance lifecycle as it requires accurate categorization and assessment of claims while sifting through massive amounts of data. Needless to say, the quality of data is of paramount importance while adjudicating claims in order to support fast and accurate claims processing.

Why It Matters

Claims adjudication is inherently a complex process because of the multitude of parties and data points involved; lack of quality data can lead to delays and ineffective management of critical business processes leading to sub-optimal results.

One of the most critical aspects that get impacted as a result of poor quality data is the speed of adjudication (cycle time), resulting in customer satisfaction (NPS) suffering immeasurably.

Inaccurate decision making can also have a significant impact on overall financial metrics (combined ratio, reserving) of a carrier. The net effect is that genuine claims take a lot longer, fraudulent claims are harder to catch and subrogation recovery is sub-optimal.

According to a recent survey of insurance companies, only 57% of participants felt they were leveraging their data analytics solutions effectively; nearly two-thirds (66%) of them expressed that data quality is the biggest challenge their data analytics program faces.

Data Quality Conundrum

There are many reasons for the accumulation of low-quality claims data, most of them arise from the way the data is collected through-out the lifecycle (at FNOL and subsequently) and maintained. Here are some of the common data quality issues and their reasons:

Inaccurate or incomplete contact information can result from duplicate customer or claimant records across various systems – Leading to difficulty in collecting information, verifying coverages and providing timely updates and payment. Data stored in individual spreadsheets, lack of common data models, data structures and data definitions – Information vacuum occurs, data sharing becomes challenging, resulting in inaccurate adjudication, leading to more expenses, potentially bigger losses and dissatisfied customers. Unstructured data stored in disparate systems, with varying degrees of granularity – Collating such information is challenging given that related metadata is usually missing or inaccurate impacting estimates, vendor recommendations and delays in accurate payments. Inappropriate labeling of injury codes and use of inconsistent abbreviations in the notes – Leading to inaccurate decision making causing delays. Data Input issues in Individual’s (customer, driver, participant, claimant) data e.g. telephone numbers, addresses, IDs, names, etc. – Inability to correlate information from various systems leads to delays in collecting information from various parties and adjudicating claims accurately. The Business Impact

As we articulated earlier, poor quality data can have a significant negative impact on accurate claims adjudication. The impact this creates for a carrier is disproportionately large.

Inaccurate Claims reserving: Accurate claims reserving is a critical piece of the puzzle for any carrier to manage their finances and investments more effectively. Regulations in most countries require carriers to have enough cash reserves to cover losses; striking a balance between accurate cash reserves and investment capital becomes critical for profitability. Inaccurate reserving directly impacts investment reserves (if reserving is too high) or worst yet creates regulatory compliance issues (if reserving is too low).

Extended Cycle times: Typical cycle times range from a few days to a few weeks. A 10-20% increase in cycle time as a result of poor data quality not only compromises customer experience and consequently customer retention; but also results in increased Loss Adjustment Expenses (LAE) directly impacting the expense ratio.

Hard to identify Fraud: According to Coalition Against Fraud, 10%-20% of all claims payments are made for fraudulent claims, directly impacting the loss ratio of the insurer. Poor data quality can result in this percentage being higher by as much as 2-3% ($ figures could be in several millions for a typical mid-size carrier)

Sub-optimal Subrogation recovery: Subrogation recovery directly impacts the loss ratio; poor quality data can lead to difficulties in identifying subrogation opportunities and clearly figuring out responsible parties based on incident details. A $2000 lost subrogation opportunity directly relates to $2000 more in losses for a carrier.

The financial impact alone on enterprises due to poor quality data is far-reaching and deep; In 2017, Gartner estimated the average cost of poor quality data for any enterprise to be $15 million. In our view, the averages for insurers are significantly higher given the data centric nature of the business, as goods and services by insurance carriers are primarily data products. Given the size of the business impact, there’s increasing awareness within insurance companies to have processes and systems in place to tackle the data quality challenges.

In part two of this blog, we will discuss measures to improve data quality specifically for claims and also look at ways to leverage this data more effectively.

Read more blogs here https://insuranalytics.ai/blog/

[insuranalytics-blog]

Adoption of AI for Auto Claims Adjudication

Insurance claims adjudication has come a long way since analytics was first used to identify authentic and straight-through claims a few years ago. Nevertheless, the auto insurance industry faces many challenges in adjudicating claims today as we enter the new decade; key among them is the rapidly rising Loss Adjustment Expenses (LAE), given the trend of increased severity and frequency of the incidents.

There is an increased urgency among industry experts to find a solution in order to control expenses and improve Net Promoter Score (NPS). The industry is increasingly looking at AI as a possible technology enabler; as AI has already been making a significant strides in underwriting, risk scoring and marketing. While claims processing has been relatively slow in AI adoption, all indications show that this will change significantly in the coming years.

Unique Challenges The auto claims adjudication process faces many challenges like delayed reporting, longer cycle times, human error in assessment or filing, fraudulent claims, customer dissatisfaction, and a lack of transparency in the process from the customers’ viewpoint.

Each step in the lifecycle of a claim after FNOL (First Notice of Loss) can take days or weeks to process. The effort required by adjusters in triaging and investigating various aspects of a claim requires significant data crunching and analysis. This is especially true for:

Checking the authenticity of a claim Accurately assessing the damages Both of the above factors are key to adjudicating the claim accurately, efficiently and quickly, with a direct impact on expenses and Net Promoter Score (NPS).

Let’s look at how an AI system can enable adjusters and the overall adjudication process to be more effective:

Checking the authenticity of the claims: Right at FNOL, the AI-powered solution can compare the claims data points with historical data and run it against known fraud markers. The solution can assess the damage, check the incident report, verify data against the policy terms, and run fraud detection algorithms to identify the nature of the claim accurately. The claim is then assigned a score and its propensity of being genuine or fraudulent is determined. The adjusters can then simply verify/validate the findings of the score without extensive triaging and take appropriate actions. Complex or potentially fraudulent claims can be routed more quickly to the experienced adjusters or SIU investigators, depending on the set parameters. Whereas, simple claims can be queued, up to be auto-adjudicated after verification.

Accurate assessment of damages: One of the most time-consuming activities in the whole lifecycle of a claim is the accurate estimation of damages, this not only impacts the final payout, but critically important for an accurate reserving amount.

AI models, trained on images of wrecked cars and incident reports, can quickly produce a fairly accurate draft estimate, which can then be used for reserving and validated against an estimate from a body shop to arrive at an accurate and verified final estimate.

As real-time image recognition becomes more effective and less expensive, it may soon be possible to offer even more sophisticated services like on-the-spot image capture, auto-trigger of towing and repair services, a transparent incident report including all relevant data, and an estimate of the settlement amount right at FNOL stage.

Enhancing customer experience and improving customer satisfaction scores: Surveys show that while one in five consumers do not like to answer too many questions at First Notice of Loss (FNOL) and prefer the self-service claim options. But lack of knowledge of insurance policy terms and inadequate coverage add to the delays and fuel even more dissatisfaction.

AI models can play a significant role in both highlighting relevant insights about a claim quickly, but also generate customer alerts based on those insights. With an accurate, transparent and timely reporting process driven by AI, the customer feels more confident about the fairness of the claims process and are more likely to accept the settlement offer. The settlement amount would be in line with the right parameters, and hence customers’ interests are safeguarded, and human errors and biases are eliminated.

The Future of Claims Settlement: Next Level of Automation Going a step further from automated claims report filing and incident reporting, another advantage that AI can deliver is automated claims support. While new customers or complex cases may require a human touch, straight-through and other less serious cases may benefit from AI-based chatbots that can coordinate the entire process and keep the customer updated about the status of their claim.

Such an automated system of claims support is not heavy on resources and is valuable for delivering much superior customer experience. While 100% touchless claims may not be possible today, AI-powered chatbots can definitely free up resources and reduce human errors to a high degree.

To sum up, there is tremendous potential to deliver a seamless insurance claims, customer experience through AI technologies; it not only helps improve Net Promoter Score (NPS), but also helps insurers reduce their Loss Adjustment Expenses (LAE) and improve cycle times, creating a win-win for all parties involved.

Read more blogs at https://insuranalytics.ai/blog/

Explore Claims AI Cloud at https://insuranalytics.ai/claims-ai-cloud/

[insuranalytics-blog]

Vision 2020: InsurTech and The Insurance Industry

The last decade saw a huge change in the way insurance industry functions. As we step into 2020 with AI-driven InsurTech initiatives, the next one promises to be even more of an adventure. Innovation and technology are going to be the forefront, but any predictions about future trends may have a tendency to be short-lived.

At the beginning of the previous decade, it would have been impossible to imagine that the consumers would trust their hard-earned to anyone other than a qualified agent, who could deliver a personal touch and inspire confidence in their investment. Today, insurance customers are willing to buy insurance and take financial advice from AI virtual assistants. If this is not a leap in the way insurance is sold, what is?

The Appeal of Gaining the Competitive Edge It is an understatement to say that the insurance business is competitive and driven by the need to acquire as much market share as possible. So it is not a stretch to see insurance company executives are already increasing their investment in technology and embracing transformation.

From KPMG International’s Global CEO Outlook:

“For insurance CEO’s, the insurtech agenda is a top priority and in many cases, at the forefront of their strategy. In fact, 73 percent of insurance CEOs agreed that they are personally prepared to lead the organization through radical transformation to remain competitive.” The insurance industry landscape is slated to change with new offerings and alliances. Insurance companies usually prefer to control their data, which may be one of the reasons they may be looking to acquire InsurTech companies, so they can deal with data regulations better and keep their data insights and products in house.

As adoption of digital tools and solutions increases, the focus on delivering an experience to customers would also increase, as delivering exemplary customer service is the way to retain and gain consumers.

The Focus is on the Customer Insurance companies are still looking to satisfactorily cater to the customers who are quite technologically savvy and are used to the ease and convenience of shopping online that allows self-service. At the same time, human interaction is not something that can be replaced, as more complex insurance products would need the expertise of agents and brokers.

Similarly, while settling claims can be automated and classified quicker and more accurately with the help of AI, the domain knowledge to train the AI to know the difference would have to come from the experts. Better case classification and settlement recommendations lead to quicker settlements and thereby increase customer satisfaction scores for the company.

The hybrid option of offering customers multiple channels – self-service or via agents and brokers – to choose products, find product information, and file claims are going to be a more popular choice. To make this kind of offering a reality, the online channels need to be designed with advanced AI capabilities to offer better advice via chatbots, and the agents and brokers need to be armed with data insights gleaned with AI-powered analytics.

Data can be leveraged to connect dots in customers’ profile to understand and anticipate their needs and requirements. Such insights can help the company cross-sell and up-sell their products, and even offer custom products to their customers on the basis of aspects like demography or geography. With the customers’ help, it is now possible to reduce the time required to choose the right product and calculate price points.

Innovate, Innovate, Innovate InsurTech companies are already focusing on delivering end-to-end solutions to bring all existing and emerging technologies together. This will be crucial in the next two to three years as insurance companies are wary of tech solutions that may not serve them in the long run and require them to purchase additional solutions as new technologies enter the market.

Technology disruption may have caused the industry to change operational ways, but adoption is still witnessing challenges. Developing customized solutions that can drive transformational change initiatives and also deliver AI advantages without black-box limitations. To aid this transformation drive, InsurTech companies will have to focus on innovations more than ever before, which may lead some to play safe and offer boxed-up solutions, whereas some may be more inclined to take risks. However, risk is something that insurance companies always look to mitigate.

The answer may lie in conducting data experiments that involve both data experts and data scientists.

This is how Jenny Rone, the assistant inspector general for data science at USDA, has applied data and domain experts:

“Most of what we’ve done focused on self-service business intelligence tools. For instance, when we received funds to oversee the 2017 disaster season, we built maps with geospatial information systems (GIS) tools. We took GIS coordinates from FEMA and we created the base based on those disaster declarations.

We then took contract data from Federal Procurement Data System (FPDS) and overlaid that based on the national interest action code. Our auditors and investigators could see the big picture and they were able to drill down on specific areas and individual contracts, which aided objectively scoping our oversight activity.”

Such initiatives are the way forward toward bringing data analytics and people together.

A Roadmap for Insurance Claims Settlement Processes While everyone is talking about adopting more AI-based solutions, there not enough focus on the implementation and execution aspect of it. The claims managers are the ones who will end up using the technology most.

Some claims managers may be required to oversee the AI data training and ensure that the knowledge from adjusters is transferred without biases into the neural networks of the AI solution. They would also need to constantly prioritize information based on each case’s requirement and ensure the criteria for straight-through processes is carried out correctly.

Hence, InsurTech companies will do well to ensure they have pilot programs designed with the claims managers in mind. They would need to:

Understand how the solution works Be able to identify inefficient processes Assess the impact of the changes on the staff Know how to evaluate pilot program’s efficiency Process feedback received from employees, policyholders, brokers and agents Claim settlement officers would also need to understand that the AI technology is here to make the process more efficient and make their jobs easier. This understanding is also possible when the pilots are executed efficiently.

While there is a lot to look forward to in the insurance AI solutions domain, there are many challenges waiting to crop up as well. The increase in the number of natural disasters and climbing figures of accidents makes the job of insurance companies and their settlement departments that much more difficult. Embracing AI technology to leverage data, both existing and new, makes sense on all levels, and when it is applied to complement human intelligence and compassion, it ends up delivering a winning outcome for everyone.

Click here to read more blogs https://insuranalytics.ai/blog/

[insuranalytics-blog]

How AI is Helping Claims Adjusters Become More Efficient

https://insuranalytics.ai/blog/

Insurance claims adjusters face a lot of challenges as they execute their responsibilities for verifying and accurately documenting the claims. The experienced ones are often over-burdened with the caseload, making the settlement cycle longer. Such delays could lead to dissatisfaction among customers. On the other hand, some claims adjusters have a lighter load and would have been able to clear more claims cases had they been assigned to them.

Claims adjusters have to check the authenticity of each claim to ensure that there is no fraud. After all, fraud is one of the biggest drains on the revenue of an insurance company. Often, the authentication process requires a long and detailed investigation, which further delays the claims settlement process.

The key to reducing cycle times is to identify the claims cases that can be settled with historical data and establish patterns that can determine STP (Straight-Through Processing) cases. This way, only the claims that need authentication to get passed on to the claims adjusters and their expertise is utilized more effectively. AI-driven claims settlement solutions offer many advantages and can help claims adjusters be more efficient.

Better classification of claims: Identifying and assigning claims is where an AI-driven claims settlement solution can help tremendously. An AI solution designed with advanced machine learning capabilities can get trained on insurance claims data and help speed up the claims settlement process. The algorithms can help sort complex cases and assign them to adjusters who have matching expertise. By analyzing the patterns from the historical data, an AI-powered claims solution allows for faster authentication, leading to the payout of genuine claims without delay.

Better fraud detention due to central access to all relevant incident data: Claims adjusters are always looking to establish the validity of the claim quickly. With all the incident data available centrally, claims adjusters are able to make the right calls quickly and with increased accuracy. Deep learning capabilities of the solution bring together all the pertinent incident data to one place. This action allows claims adjusters to choose the claims that need investigation and those that need to be deferred to the SIU (Special Investigative Unit). Better fraud detection also directly results in an improved loss ratio.

Better customer experience due to faster settlements and accurate estimates: With faster and more accurate claims processing, cycle times become shorter. Claims adjusters can use predictive insights for faster closure of genuine claims. This ability leads to a reduction in expenses, which, in turn, delivers better NPS (Net Promoter Score) for the insurers. Better customer experience also leads to better customer retention.

With AI claims solution, claim adjusters can also help identify patterns from data to catch repeat offenders and fraud networks.

Settling insurance claims is a complex process, and claims adjusters are a vital part of it. They are responsible for authenticating the claims and make sense of all the incident data to arrive at an estimate. They are overworked and under a lot of pressure to settle claims quickly. At the same time, they have to ensure that no fraudulent claims are paid out. AI-enabled claims settlement solutions can make their job easier and help them streamline their claims caseloads.

Click to read more blogs https://insuranalytics.ai/blog/

[insuranalytics-blog]

3 Top Ways in Which AI Can Minimize BI Claims Leakage

BI (bodily injury) claims leakage is a major drain on an insurance company’s resources, mainly because it is difficult to identify all the symptoms of the injury caused only due to the collision or occurrence. Since not all symptoms can be identified during the initial assessment, triaging these claims consumes more time when compared to other types of claims. Even experienced adjusters cannot accurately predict and reserve the payouts needed to handle BI claims.

WHY THE BI CLAIMS ARE SO COMPLEX Too many factors to compute for payout: Most of the BI claims are high payout claims due to the severity of the injury to the claimants. The volatile nature of BI claims settlement makes it very difficult to reserve for BI exposures, due to the range of factors that need to be assessed for reaching a payout figure such as age, gender of the insured, type of the collision, position of the insured during the collision, any known or unknown pre-existing health conditions, and several other aspects.

Multiple parties involved: To make matters more complex, most of the BI claims involve multiple parties – insured and third-parties, leading to an increase in time required for settlement. These delays in settlement cause customer dissatisfaction and failure to quickly and correctly identify the nature of claim triggers an increase in exposure to litigation.

More time required to process claims: BI claims require more time to process as these type of exposures require not only an initial assessment, but also several sessions of treatments and continuous monitoring by medical professionals. A specialist insurance adjuster is needed to confirm that the required injury is handled and the insured or claimant is rehabilitated.

HOW TO TACKLE THE PROBLEM OF BI LEAKAGE BY LEVERAGING AI 1. Get the data

Apply AI models to catalogue incident details: Artificial Intelligence can be applied using various models like Random Forest, Bayesian Networks, or a combination of multiple models, to scan through claim notes and analyze images or videos of the accident to catalog details of damages and the injuries as the claim progresses, and provide intelligent predictions.

Leverage historical data: Compare the current claim’s FNOL and initial assessment notes to historical records and draw up a list of potential early predictions of all the symptoms and hidden or derivative injuries that could crop up at a later stage.

2. Streamline the process

Assign the right adjuster: By having the claim assigned to the right adjusters’ group during the initial stages of the claim and as the claim progresses, insurers can look forward to timely triaging of the claim.

Manage vendors efficiently: With the ability to identify the symptoms during the initial stages of the claim, insurers can now make early referrals to the right medical vendors and rehab facilities to take care of the injuries faster.

3. Predict payouts accurately

Predict payments: AI models can provide consistent reserves and payment estimates for the respective BI exposures with similar collision conditions, location, age, gender, position and number of claimants involved, the severity of the collision.

Make an early settlement offers: With early insights, claims adjusters would be able to negotiate with the claimants by providing early lump-sum payment options, which is a desirable outcome for the claimants, third parties, and also the insurance organization. Claimants and third parties get their much-needed payout early to settle expenses, and the insurance company can minimize its exposure to litigation and reduce costs by closing the claim fast.

HOW AI HANDLES THE COMPLEXITY OF BI CLAIMS AND REDUCES THE IMPACT ON INSURERS Disadvantages the insurers face due to BI claims complexity:

Susceptibility to soft fraud: Chances of soft fraud by a claimant or third parties increase for the claims involving bodily injuries since it is extremely difficult to prove whether the symptoms of the claimant are due to a pre-existing health condition or due to the injury. As BI payouts are generally higher than other types of claims, not being able to identify potential frauds adds to the insurer’s losses. Potential exposure to litigation: The insurers are more vulnerable to lawsuits involving BI, as early detection of the symptoms and impact is often missed. In several BI claims, some of the key symptoms might only crop up at later stages due to the inability of the claims processing team to identify all the underlying symptoms of the damage during the initial assessment. In such cases, the insurance company becomes exposed to litigation on negligent grounds. Negative impact on the bottom line: Insurance companies end up paying more for settling BI claims due to process inefficiencies and failing to detect underlying symptoms at an early stage, which negatively affects their bottom line. Benefits due to the application of AI:

Reduced processing time: The AI solution is capable of going through thousands of claim notes and provides key insights from prior claims quickly and accurately, which is otherwise very time consuming even for experienced adjusters, and absolute accuracy is never guaranteed due to the human factor. Increased customer satisfaction: Taking care of the claimant’s monetary needs faster increases the claimant’s satisfaction, which helps with the NPS. Further, since the overall claim processing time and claim payments are reduced, this provides the insurance organization with the opportunity to reduce the premium amounts, which also increases the insured’s delight and works toward earning their loyalty. Reduced or optimized payments: The ability to predict and handle injuries and hidden symptoms at an early stage before they get aggravated, helps with faster healing and recuperation of the claimant, which can significantly avoid additional payments for the treatment of hidden symptoms. Settling claims is a resource-intensive process that requires great attention and care, and cost insurers a significant part of their revenue. The more complex a claim, the more resources and time it requires for settlement. With better categorization of injury data, it is also possible to authenticate and indicate red flags more accurately.

By leveraging AI technologies, it is possible to reduce significant resource drain for most insurance companies that occurs due to BI claims leakage. A neural network trained on robust data offers better pattern recognition and delivers more actionable insights.

Read more blogs https://insuranalytics.ai/blog/

[insuranalytics-blog]

Impact Of Computer Vision In Insurance - InsurAnalytics

Almost 95% of insurers still rely on tabular and text data for #ML models for pattern analysis. With advancements in deep learning, computer vision applications are now able to leverage unstructured and image data to solve many insurance claims related challenges. Murugesan Vadivel explores in detail.

[insuranalytics-blog]

AI In Claims Litigation

https://insuranalytics.ai/blog/

A claim litigation process is started when a claimant or third-party insurer is not satisfied with the claim decision or the payment offer. Insurers have to find ways to defend themselves against the lawsuits from their policyholders and third party.

The general steps involved in handling litigated claims are:

The exhaustive nature of claims litigation process makes it very time consuming and costly. Even though very few claims go to litigation, the claims settling cost reported in insurers’ financial statements as “Defense and Cost Containment Expenses (DCCE)” are quite high. The following figure shows various costs that sum up as DCCE.

The graph shows DCCE expenses for five consecutive years from 2013-2017, depicting a growing trend. Defense and Cost Containment Payments incurred only for private passenger auto liability/medical by insurers as a percentage of total payments made in the year 2017 is 4.21% as per the data provided by The National Association of Insurance Commissioners (NAIC).1

Costs of defending certain types of lawsuits, such as medical injury cases and class actions against pharmaceutical companies are relatively high. This is reflected in the annual financial statement of the insurance market released by the Insurance Information Institute (III). As per the financial statement in 2017, in addition to $940 million incurred in product liability losses, insurers spent $648 million on settlement expenses, equivalent to 68.9% of the losses.2

The primary reason for losses incurred by insurers as DCCE is the difficulty in predicting a lawsuit in the initial stages of the claims process. Hence, the claims representative has to thoroughly scan through all aspects of the claim details relating to the incident, plan & process each claim in good faith, to ensure optimum payouts and avoid future re-opening of these claims due to potential litigations.

AI can be used to, not only, predict the probability of litigation, in the early stages of a claim, but also provide inputs to enable decision making & take necessary steps for faster & optimum resolution of claims. AI has the advantage of using large amount of diverse data sources to obtain unique information for segmentation and analysis of claims. This helps to identify the combination of complex patterns that point to a potential risk of litigation.

Employing AI, we can assess each claim & provide below predictions, in the early stages of a claim life cycle, to ensure better outcomes:

Probability of litigation in a claim

Key team members for successful outcome of litigation

Optimum payout offer for quick & satisfactory settlement

Optimum time to resolve a claim

The involvement of the right teams from the early stages of a potential claim can result in a reasonable compromise between the claimant/third party and the insurer or better preparedness to face the litigation process. Artificial intelligence algorithms are also capable of mining Big Data to find out which attorneys win which types of cases and before which judges. Hence appointing the right attorney in the right case can increase the chances of winning the litigation. AI algorithms can also be employed in obtaining invaluable insight on optimized settlements for specific historic cases through exhaustive analysis of historic claims. The high computational power associated with AI provides instant access to quality and quantity of information about similar claims; this can help take the decision on whether to fight or settle a pending lawsuit. This results in substantial savings in time and money for the insurer.

Conclusion

With the integration of AI in the claim life cycle, the risks associated with a claim going for litigation can be predicted effectively. With this as input, and involvement of the most appropriate team members suggested by the AI algorithm, the defending or settling of a claim by compromise becomes easily achievable, potentially impacting the customer experience, adjustment expenses and losses for the insurance carrier.

Citation:

https://www.actuary.org/files/publications/PC_RBC_UWFactors_10282016.pdf

https://www.iii.org/fact-statistic/facts-statistics-product-liability

Read more blogs here - https://insuranalytics.ai/blog/

[insuranalytics-blog]

Is Machine Learning and Artificial Intelligence in Insurance Showing Tangible Results?

https://insuranalytics.ai/blog/

Incorporation of artificial intelligence (AI) applications in insurance are disrupting traditional insurance practices. AI use cases are becoming more common where insurance companies have achieved better customer satisfaction. Automation has helped reduce costs for insurers, enabling them to provide better coverage for lower premiums, resulting in enhanced customer loyalty and retention.

There is still a long way to go before AI completely penetrates the insurance industry, and companies pass on the benefit accrued to their target customers. Early adopters are however reaping the maximum rewards. Some areas where the insurance industry has implemented AI to yield tangible results are:

Usage-based insurance On-demand insurance Peer-To-Peer (P2P) insurance Usage-based insurance (UBI) UBI, better known as pay as you drive (PAYD), is a tailor-made vehicle insurance policy based on the type of vehicle, distance, run type, and driving behavior of the individual.

Data acquired using telematics and mobile apps are used to determine the premium for each individual. Telematics devices connected to automobiles capture data from the onboarding kit, providing vehicle information. This, along with driving behavior and combined with GPS information, allows the insurance companies to determine customer behavior. Mobile apps that capture the customer behavior (speeding, braking, acceleration, and distraction) have been developed by combining the GPS information and phone usage while driving.

Differentiation based on individual driving behavior allows the insurance companies to offer lesser premiums for safer drivers. This leads to greater customer satisfaction and an enhanced customer base. This is indicated from a survey conducted by JD Power & Associates in 2016 which found that “…UBI participants provided more positive recommendations and more often indicated that these recommendations resulted in a friend, relative or colleague purchasing from their insurer compared with those customers who did not use a UBI program.”

Earlier, insurance companies used to gather historical claims data and geographic data and run analytics to predict the future. Presently, telematics devices, sensors, and smart phones are the devices that provide insurance companies with the data they need to better assess the customer as ‘risky’ or ‘safe’. Based on the driving data captured from these devices, plus the legacy data they have, a large number of behavioral models and algorithms are developed. These models give almost accurate predictions based on the datasets provided. With increase in data, the model enters a self-learning phase, i.e. it matures and provides more accurate results. In future, companies will be moving their pricing models from likely predicted pricing to actual driving behavior of the customer in real time.

On-demand insurance or pay as you go insurance On-demand insurance policies can be bought with the click of a button, whenever and wherever required, without directly interacting with a broker or a company representative. These policies include coverage for the on-demand economy (such as Uber drivers); they provide flexibility of term and pricing. Applications like Chabot have enabled the insurance companies to automate the policy buying experience of the customers. Based on the information provided by the customer, Chabot can be designed to pull social and geographic data to provide better coverages with good premiums in shorter duration. Better and faster interaction can lead to more business. These apps can be used to provide on demand or pay as you go policies. Companies like UBER, Lyft, Turo, Vehicle Rental agencies, etc., are already partnering with insurance companies to provide the drivers with on demand insurance for a specific period.

Automated platforms provide better efficiency with reduction in issuing time of policy. The various processes involved, from identifying the individual, to authenticating the credentials, fixing a quote, and binding a policy are all automated to provide better customer experience. According to a survey by Accenture, 74% of customers are willing to interact with computer generated systems for insurance advice.2 This personalized experience is more attractive because of the convenience and practically zero response time.

Peer-To-Peer (P2P) insurance Insurers are designing models to speed up claim processing while reducing frauds to make way for a better customer experience. The obvious choice to achieve faster processing speeds in real time is by creating better AI models. This can be achieved by combining social media data, geographic data, legacy claim data, data from telematics devices, and GPS data from cell phones to assess the claims in real time.

Faster claims processing and better coverages lead to better customer satisfaction and increased sales. However, to achieve this, regular analytics will not be sufficient. Self-learning AI models that can read huge data sets from various sources in real time and generate accurate results will lead to better business outcomes.

Conclusion: The adoption of AI in the insurance industry promises to bring about the much needed change from traditional insurance practices to a more customer-and business-centric approach. Self-learning AI models capable of generating accurate predictions in real time promise to lead to better policy coverage, premium estimation and settlement.

With an increase in customer willingness to shift to smart technologies, greater adoption of AI in insurance promises tangible growth in business with reduced costs and greater customer satisfaction.

Citations: https://www.insurancejournal.com/magazines/features/2016/07/25/420531.html

https://www.accenture.com/t00010101T000000Z__w__/gb-en/_acnmedia/PDF-50/Accenture-Distribution-Marketing-Survey-Insurance-Report.pdf

Read more blogs - https://insuranalytics.ai/blog/

[insuranalytics-blog]

Top 3 Predictions in 2018 About AI in Insurance: Hits or Misses?

https://insuranalytics.ai/blog/

AI and ML (machine learning) were the big buzzwords in 2018. There were predictions regarding the increasing adoption of AI and then there. So far, AI has been adopted across various insurance processes, such as claims settlement, sales and distribution, product development and design, policy management, customer service, and even underwriting.

Here are some AI predictions from 2018 and my viewpoint on how they fared.

1. AI Will Be the Magic Wand and Bring Higher Value Than Expected The Claim:  AI and ML will start bringing in value from the day an insurance company implements it.

The Reality: Barring a few success cases, most of the AI product and platforms are still in the experimentation, evaluation & prototyping stages. Significant value expected in 2018 is still not visible / realized by the companies that have invested.

The Impact in 2019: As AI and ML will emerge from their experimenting phases to Operationalization, we will be able to see their effect on the industry & outcomes. The focus will be on proving the technology’s worth and making it even more accessible. Stakeholders will look to see the theory turn into reality.

2. Exponential Insights for Insurers from the Capabilities of AI The Claim: Machine learning will take your data and turn up miraculous insights, as soon as gets trained on your industry’s data patterns, which can be sped up with more complex algorithms.

The Reality: It turns out, training an AI algorithm takes time. Even plug and play systems need to pre-train their neural networks on industry data if they wish to show results quickly. However, this need to train neural networks quickly has given rise to innovative new solutions that are rethinking basic computing models.

The Impact in 2019: With the hype around AI diminishing, we will actually be able to witness the real value that machine learning can and will provide. Platforms with pre- trained networks will be able to get a leg up in the industry, as compared to the products with newer algorithms but no training. Reinforcement learning will play a key role. Better AI training solutions will show up and make complex insurance processes more streamlined.

3. Better Customer Satisfaction Through Efficiency in Claims Settlement The Claim: Most major insurers will adopt AI, and the consumers will finally reap the benefit in terms of hassle-free claims settlement and better customer service.

The Reality: While AI has made the process more streamlined, we are still some ways away from “touchless claims.” The drone capture of the accident site may have begun, but there’s still some time before a completely seamless AI-driven process takes over claims settlement.

The Impact in 2019: It is entirely possible that an AI platform autonomously starts taking care of claims. However, as an industry that is wary of all things new and innovative, it is possible that more regulation may arrive to govern the use of data and the role of data scientists. A transparent system that can show how data is collected, standardized, and managed, will be able to move along the adoption of AI in every step of the process much faster.

And the Burning Question! The Claim: Many jobs will be lost to AI as it moves across industries.

The Reality: The application of AI will create more jobs rather than take them away. It will mostly make redundant the cumbersome process of collecting and analyzing data, mainly those aspects that are prone to human error and are impacted by human computation limits.

The Impact in 2019: We will need more people who can think outside the box in ways to make the algorithms more robust and efficient. We need people to train AI. As data volumes increase around the globe, we will only need more people to work with it.

AI technologies are here to stay, and they are making a massive impact on the insurance industry and its various value chains. The effect may take a while to register, but it is sure to expand exponentially. The key to staying relevant and not getting drawn into the hype is to focus on your needs and find the right AI solutions that can deliver it without adding to your cost of ownership.

Read more blogs at https://insuranalytics.ai/blog/

[insuranalytics-blog]

Application of AI in Insurance Claims Settlement: What’s Happening and What’s Next

https://insuranalytics.ai/blog/

Industries all over the world are widely adopting AI technologies and the insurance industry has a bit of catching up to do. However, there is one area of insurance where AI has made a massive impact, and that is insurance claims settlement. I went as a speaker at the Pune Data Conference 2019 where a bunch of us from various companies gathered to exchange ideas and experiences about the way AI’s application is growing in insurance and how to handle this technological revolution.

Here are a few points that I discussed at the conference. I opened by discussing the challenges claims processing, most of which you may already be aware of:

Claims settlement is a time-consuming process due to the various levels of authentication and verification required before the claims payment can be released. Claims processing is error-prone, mainly due to the human factor. The cost of settling claims is upwards of 30% of the revenue of an insurance company. Insurance fraud is a stark reality, and the insurers and their customers lose billions of dollars to it every year. The current estimate is at USD 40 billion for non-health insurance. Customer dissatisfaction is a direct result of an inefficient claims settlement process. Add to this the cost of fraud that raises premiums, and you have customers ready to add to the churn. With data getting digitized, the threat to the security of customer data also increased. Every day, insurance companies fend off an increasing level of cyberattacks. Protecting sensitive customer data and information is critical and also expensive. How AI can Help Mitigate the Challenges AI solutions can fast track claims by efficiently segmenting them to enable quick and accurate claims settlement and reduce claim adjustment expenses. By evaluating various data parameters, it is possible to detect an anomaly in claims on the basis of customer behavior and reduce fraud. Getting the right adjuster assigned to a particular claim is critical for settling it with the correct payout promptly. Insurers can also use AI solutions to predict the subrogation potential of a claim. As customer churn is a huge challenge for the industry, using AI for assessing customer behavior and use it to reduce the churn rate. How We Developed the AI Models to Tackle Different Use Cases We determined that a sprint-based plan was most suitable for AI Model build and validation. Typical sprints for each use case looks like below:

A Typical Sprint

From here, I dove deeper into explaining how they worked towards managing the biggest challenge in training neural networks – feature selection. It is imperative for a neural network to know and understand which data points it needs to consider and which are irrelevant.

By using a systematic data profiling program, we were able to bring down 10000+ data points to 100+ relevant ones.

I discussed the series of experiments we did to find out the best way to train the neural networks which included working with CNN, RNN (LSTM & GRU), Transformer, Attention, ELMo, BERT, and Word Embeddings for unstructured text and language modeling.

As the discussion moved towards machine learning, it became obvious that looking into the AI black box is also becoming important. The end-users of AI technology are increasingly looking to understand exactly how an algorithm reaches a conclusion. This understanding is necessary to ensure that the neural network is getting trained properly and is using the correct parameters and relevant data points. I demonstrated how one could look into the model learnings during training using TensorBoard and other model visualization tools.

From here, the challenge of unstructured data took center stage, and I was able to demonstrate the results that we have been able to garner by using deep learning models. In claims settlement, much information is recorded as unstructured data, such as photographs of the damaged vehicle, incident reports and receipts, emails between various parties, and most importantly, notes written by the adjuster in shorthand.

We further explored attention-based models and their significance in an interactive session. The point that emerged was that designing a proper multi-headed self-attention mechanism is one of the most effective methods of training sequence-based models. It also helps us with our unstructured data points.

We also went over the latest trends in the NLP overview. Ideas about applying Transfer Learning in NLP models were discussed, along with its past limitations that can now be overcome. The latest research coming out of Google and OpenAI in Deep Learning space was also a hot topic of discussion.

AI is slated to play a critical role in automation and improvement of daily insurer operations involving image and text processing. If it can be successfully demonstrated that an AI model can offer massive benefits to insurers such as reducing current costs, improving outcomes and thereby reducing future costs, and keeping their customers happy, the insurance industry may be more inclined to adopt the technology, even though it is a highly risk-averse industry.

Read more blogs https://insuranalytics.ai/blog/

[insuranalytics-blog]

AI and Insurance: Why Domain Expertise Will Matter Even More in 2019

Solution architect Varun Chutani writes why AI applications need support in the process flow, algorithms, and system checks from insurance domain experts. Read more from his blog.

Some studies are showing that investment in AI has already slowed and that companies are looking to verify the promised ROI of deployed AI technologies before taking their adoption drive further. At the same time, with increased understanding about AI’s role and capabilities, it looks like this year insurance companies will choose those applications of AI that matter to their business and see past the just the shine and glitter of AI.

To ensure that the AI solutions deliver both business value and customer satisfaction the industry leverages domain experts need to train the solutions and develop additional parameters for its continued learning.

[insuranalytics-blog]

Next Step in Adoption of AI in Claims Processing: Challenges & Solutions - InsurAnalytics

Claims processing cost and fraudulent claims payment can increase the operating cost of insurance companies and greatly inconvenience customers.

Adoption of AI in claims process results in consistency in handling of claims, reduction of turnaround time, and accuracy in claim settlement for both the insurer and the policyholder. It also reduces the operating cost and resource use of the insurer. However, there are several challenges that need to be overcome before AI can be an integral part of claims processing. Strategic planning towards AI adoption with a solution based approach to minimise risks involved in handling of large data sets, extensive testing and training, and ensuring regulatory compliance, can help the insurance industry to process claims at a lower cost and achieve better business outcomes.

[insuranalytics-blog]

AI-driven Claims Processes Can Ensure Smoother Settlement During Natural Disasters - InsurAnalytics

AI-driven claims processes can ensure smoother settlement during natural disasters.

AI technologies like machine learning and deep learning are already being used to in claims settlement processes and these AI-driven solutions have already started delivering results. It’s now time to embrace the AI and deep learning technology at a much broader and faster pace that can help make the lives better for the consumers. When the next disaster strikes, the company will be prepared for faster and cheaper claims settlement process and would be one less thing for everyone to worry about.

[insuranalytics-blog]

Deploying Deep Learning in Claims Reserving - Read Our Latest Blog

Learn Deep Learning can improve Claims Reserving with data models and pattern-based analytics & how best to avoid chances of over-reserving or under-reserving claims.

Deep Learning has the capability to offer multiple opportunities for P&C insurers to the better augment, particularly in Claims reserving. Data that can be omitted or overlooked through human error can be analyzed in detail using the tools of artificial intelligence and help reduce costs/liability for the insurance company. Without ignoring the traditional reserving methods, Claims reserving process will definitely be augmented using AI methods. Modern Technology would be the major factor influencing this change.

[insuranalytics-blog]

Leveraging AI to Mitigate Fraud Payouts - InsurAnalytics

The insurance industry is undergoing a transformation by an increasingly digital world. Customers expect faster claim processing and payouts while insurance companies are looking at ways to reduce the cost of processing claims.

The insurance industry is undergoing a transformation by an increasingly digital world. Customers expect faster claim processing and payouts while insurance companies are looking at ways to reduce the cost of processing claims. Using AI to automate various parts of the claims process is a step in the right direction to achieve both these goals. While undertaking this journey, it is important that fraud detection not take a backseat.

[insuranalytics-blog]

Predicting Subrogation Potential - Role of Deep Learning (Part 2) - InsurAnalytics

In part 1 of this blog, we discussed why subrogation is important and how effective subrogation positively impacts the company’s bottom...

In part 1 of this blog, we discussed why subrogation is important and how effective subrogation positively impacts the company’s bottom line. We also enlisted the many causes why potential subrogation opportunities are missed. These could be manual methods of evaluation, lengthy and costly investigation processes, and complex rules governing subrogation. These drawbacks are compounded due to overworked and understaffed teams who are not able to effectively prove third party liability and collect subrogation dues. In the following blog, I will discuss how predicting subrogation through deep learning helps to effectively reduce losses and provides greater customer satisfaction and enhanced loyalty.