Transforming the Claims Process with AI
April 09, 2021
The insurance policy is a contractual agreement that stipulates, in exchange for payment, a person or business will receive reimbursement or financial protection against losses from an insurance company. In other words, if the required premium is paid, a customer will be paid for covered damage, injury, illness or loss.
The insurance claim is the formal request for coverage for a loss or event.
For as many kinds of insurance that exist (from workers compensation to travel to life), there is the potential of there being a corresponding type of claim. And while the details of a claim for short-term disability benefits and a claim for auto damage following an accident are different, you might be surprised to hear that the overall process flow is similar.
When thinking about customer interactions and processes, we find it helpful to first think about claims as either those that physically affect people (i.e., injury, illness or end of life), those that impact physical property (e.g. vehicles, buildings, equipment) or those that involve service (e.g. travel, events).
For any claim, there is a claim intake process, where the customer makes their formal request for coverage under their policy. Insurance carriers have different terms for this step, including First Notice of Loss (FNOL), or Filing A Claim. Sometimes a claim will be reported by someone other than the policyholder – for example, the owner of a car that was damaged in an accident where they believe the policyholder is responsible.
Policy verification involves confirming that there is a policy in effect that matches the facts within the claim. That the customer has a policy in effect on the date of event or loss is key. Other elements of policy verification will vary based on the type of claim. For example, for a short-term disability claim against a group benefits policy, the injured party needs to be verified as being a member of the group. For a personal auto claim, the car needs to match what is shown on the policy.
Claim assignment needs to happen both as quickly and accurately as possible. Claims adjusters may specialize not only on a specific line of business (say, workers compensation) but may even be focused on specific industries or jurisdictions. Business customers may have dedicated claims handling teams. Sometimes claim specialist assignments are based on reported loss severity. Regardless of the claim assignment rules in play, the insurance company wants every claim to get into the right hands to provide the best and right-fit service immediately.
During the claim investigation the claims adjuster obtains the facts needed to confirm coverage – that the claim is in fact covered under the terms of the policy. Policies have exclusions that limit what is covered. This can be to prevent overlaps between types of insurance, like homeowners and flood or workers compensation and disability. The claims adjuster also confirms the cause and extent of the damage or injury, where there may be applicable terms in the policy. This could involve obtaining a report from a doctor or obtaining a repair estimate from a contractor. The claims adjuster looks for other responsible parties, like the manufacturer of a defective part that caused a machine to malfunction, should there be a reimbursement claim to pursue, referred to as subrogation.
Some claims will have a longer duration, including significant building losses or injuries. During the claim management process, the claims adjuster is in regular contact with the customer or injured employee. For disability or workers compensation claims, the adjuster monitors medical treatment and return to work readiness. For a larger property loss, the adjuster is following the progress of repairs. And in all instances, the claims adjuster is ensuring any payments due are being made timely.
Claims eventually reach the point of claim resolution, where all payments have been issued. At that point the claim file is closed or moved into a recovery process if there is subrogation potential.
The claim journey is highly interactive, from intake through resolution. As outlined above, most claims follow a similar flow of information received, processed and requested. This creates many opportunities to leverage AI to automate and accelerate these anticipated touchpoints. Each discreet customer, agent, broker, member, claimant and even provider interaction that is automated, can simultaneously improve customer experience, operational efficiency and employee engagement.
We offer just a few transformation possibilities here with the hope they will inspire you to reflect on your own high value opportunities.
The claim intake process sometimes begins with a customer or their agent emailing a form to their insurance company. An integrated AI platform that includes machine learning and email automation can rapidly recognize new claim forms as they land, extracting the data and moving it into the core claims system in minutes. This eliminates hours – if not days – of delays in policy verification and claim assignment, and in turn, customer service. Absent that automation, carriers have to deploy vital employee resources to monitor email boxes to rekey each new claim, which can hinder the claims adjuster’s ability to manage each claim to its best outcome.
During the claim intake process, AI can anticipate – and ask for - the information that will be needed from the customer as part of the claim investigation. This could include facilitating the e-sign of a medical authorization and introducing them to in-network medical providers specializing in their type of injury. Or AI can enable them to schedule an appointment at a preferred repair facility nearby. A health plan’s need to coordinate benefits before being able to pay a claim can be easily addressed by a member uploading a photo of another secondary plan card.
Throughout the claim investigation, management and resolution phases, AI empowers carriers to be proactive, automating the answering of customer questions before they are even asked. Claim status is a high-volume driver of inbound customer calls and emails for most carriers. The customer, member or claimant experience is transformed when they automatically receive regular updates, from the progress of their appraisal, to the timing of their next benefit payment or confirmation their medical bill has been paid.
Insurance companies can also leverage AI to automate reaching out when information is needed, be it asking the customer for the location of a vehicle to be appraised, checking in with a member for the timing of their next doctor’s appointment or confirming the desired payment process for a life insurance claim.
Insurance carriers formed data analytics teams several years ago, to tap into the power of decades of loss data using modern desktop business intelligence tools. This enabled carriers to gain new insights, including patterns of claim outcomes. Predictive models followed, where insurance companies used historical trends to predict new claim outcomes. This not only improved the accuracy of loss reserves (think claim payout budget), but it also helped carriers to see early indicators of initially simple claims that historically became complex. However, it wasn’t an integrated process – claims handlers were provided lists of claims where intervention was needed.
Machine learning within an integrated AI platform enables carriers to insert the power of data analytics into the claim process to automate both predictions and decisions. Utilizing historical claim data, ML can find the early indicators of fraud and automatically notify a Special Investigation Unit. Injuries and illnesses that have a strong likelihood of worsening could be routed to a nurse to arrange a medical second opinion. Losses identified as high contenders for future litigation might be elevated for manager review and reserves adjustment. These are just a few examples of proactive automation that can both effectively and efficiently improve claim outcomes.
Insurance customers are also consumers whose automation expectations have been heavily influenced by digital leaders like Amazon, Netflix and Facebook. Gartner’s research indicates customers expect to manage 85% of their relationships with businesses without interacting with a human. According to Microsoft, most customers of any age will select a self-service channel first.
Integrated conversational AI platforms enable carriers to meet their customers where they are, in their digital channel of choice or to connect them to the right person. This not only positions carriers to offer the self-service options customers expect. It also enables carriers to automate high volume customer outreach, including differentiating touchpoints beyond the traditional policy, billing and claims process.
No-code platforms put creating new workflow automations in the hands of business analysts. These “citizen developers” position carriers to continuously introduce new solutions to respond to an increasingly digital world.
Connected devices including wearables are already used by almost 40% of insurance customers today and more than 70% expect to be users in the future, according to Bain & Company. McKinsey estimates there will be up to one trillion connected devices by 2025. Connected devices create customer value in preventing claims and can offer an insurance carrier valuable data for underwriting. A surprising benefit identified by Bain & Company is that connected devices drive customer loyalty, increasing customer-carrier interactions five-fold, in an industry where 1/3 of consumers have no interaction with their insurance company or health plan throughout the year.
Robotics will continue to permeate our everyday lives. There will be more use of 3-D printing in manufacturing and construction. Not only will more cars gain autonomous capabilities, like self-driving. Similar growth will occur with farm equipment, programmable drones and surgical equipment. These waves of change will require insurance products to keep pace, including designing different ways to understand and price new risks. Claims handling will require new capabilities and resources to appraise damage, determine coverage and liability. When a car is involved in an accident while operating in full-autonomous mode, a software engineer may need to be on call to evaluate computer code to determine responsibility.
