For years, insurance companies have used external data, in or near real time, combined with internal data and powerful analytics to help them evaluate and determine the appropriate tier and rate for policy applicants. This has allowed insurers to quickly and accurately segment risks, using the most current information available about the prospective risk, generally real time, with the customer online, or on the phone.

The leverage provided by this use of technology in underwriting has also allowed insurers to dramatically redefine and improve their processes. Not only has the overall rate-quote cycle time been dramatically reduced, which benefits the consumer and allows the insurer much higher throughput, but the quantity and quality of data obtained in the process has improved, allowing further refinement of risk and segmentation of pricing in an ongoing continuous loop of process improvement.

Claims organizations generally do not use multiple data sources and analytics engines as regular tools when triaging an initial loss report, or throughout other points in the claim lifecycle. When claims does obtain and use data outside of the claims system, it is generally point-in-time information.

This article explores the issues and opportunities present to more proactively “underwrite” claims using both internal and external data, combined with advanced analytics, to perform a more accurate initial triage and throughout the life of the claim to compress cycle time, which is a win for both customers and insurers.

The Underwriting Analogy

Personal lines companies have developed the art of rating an auto risk into a highly disciplined science by segmenting these risks into thousands of rating cells, using a multitude of factors that are calculated via complex multi-variate rating algorithms.

The insurer combines internal data with external data (some public, some private) from multiple sources, processes this data through rating algorithms to match a rate to the offered risk, and delivers back a proposed rate that accurately reflects the risk (assuming the supplied information is accurate).

Personal lines carriers have developed this rating approach into an almost real time process. Customers and/or agents can get a relatively accurate quote within minutes, even though the insurer has to obtain data from multiple sources, including external resources, to obtain information such as credit reports, vehicle history and driver history, etc.

This auto underwriting process is a fine tuned meshing of technology, data accumulation, analytics, product management skills and actuarial activity. This process is a dynamic one, which works to assure that the most recent data is used in modeling, analysis and rating.

Benefits of Underwriting Approach

With this multi-faceted approach, insurers have been able to improve their results on several levels. In addition to improving rate segmentation granularity, they have been able to take a process that was highly dependent on internal data, added current external data, combined it with advanced analytics and multi-variate analysis to produce finely segmented risk analysis and pricing.

While this underwriting transformation has benefited carriers on the accuracy front, it has also benefited carriers on the service delivery front. Carriers have leveraged technology and advanced analytics to enable them to obtain current external data (MVR, C.L.U.E.) in real time that allows them to deliver a rate quote while the customer is online or on the phone.

The consumer benefits in speed, convenience and accuracy. The insurer has significantly reduced the rate quote cycle time. The insurer—by effectively leveraging people, capital and technology—runs a more efficient operation, allowing them to pass the reduced costs on to customers in the form of lower rates and greater ease of use.

Bringing Best Practices to Claims

If insurers can use a broad swath of real time data (internal and external) combined with advanced analytics to more efficiently and effectively underwrite risks, why can't the claims area also use the same approaches to more efficiently and effectively “Underwrite Claims?”

Every insurer strives to provide claims handling that is fast, fair, accurate and customer focused. When an insurer receives a new first notice of loss (FNOL), the insurer claim system generally queries the policy system for data, matches basic insured participant data against the policy and assigns the claim to a basic handling track based on a handful of relatively static rules, generally in a decision-tree fashion.

This next generation approach to claims handling uses an underwriting-like approach, beginning with FNOL triage and takes it a step further by applying this approach of using multiple sources of data and advanced analytics to effectively re-triage the claim throughout the claim lifecycle as new and/or additional information becomes available.

Claims handling is a complex process, with multiple effected parties and multiple dynamics in play at the same time, which can impact process and progress on multiple fronts. A data pre-fill process using external data in addition to internal data will allow more accurate initial triage of the claim and exposures. For example, knowing with certainty that the claimant is insured at FNOL will assure the claim gets on the right track at a much sooner point in the lifecycle of the claim, rather than waiting until the claim is closed and sent to internal or external subrogation. Another example is using an external data fill process during FNOL to provide information relevant for CMS reporting, eliminating the steps and touches the adjuster might need later on in handling to secure data relevant to CMS reporting.

Combining multiple sources of data with advanced analytic tools would allow the claims organization to develop much more sophisticated company-specific claims triaging and assignment algorithms, both at FNOL and periodically throughout the life of the claim. This approach would allow transformational process improvements vs. simply looking to incrementally reduce the cycle times of the current processes.

In all claims departments, new/additional information is received periodically throughout the life of the claim. More proactively pushing this new or additional data into robust analytic engines can provide actionable information that allows the claim system to essentially re-triage the claim, or individual exposures on the claim, with alerts to the handling adjuster and/or manager based on company determined rules.

Execution

In the claims world, the impact of using analytics to provide actionable information can significantly enhance the leverage of people, capital and technology. Every insurer has oceans of data. Not many claims organizations have effectively turned that data into actionable information that the adjuster can use to move the claim process forward in a timely manner. Fewer yet have combined internal data with external data into analytic models to provide an even richer palate of actionable information. To be actionable, information must be relevant, timely and consumer oriented. The consumers in this context are the areas within the insurance organization that would use and derive benefit from the information, including areas outside of claims. An example might be information from an outside database that indicates the driver of the insured vehicle may not be a resident of the household.

Another current process example would be information from the claims data warehouse indicating a persistently high frequency of treatment on PIP claims with a specific medical provider across numerous claims. This data could be useful in determining if this group of claims might warrant pursuit of external data and/or referral for review by the SIU area. It is useful to actuaries in helping understand and project frequency and severity.

Information is of limited use if it is not timely. It would be critical for this information to be identified as soon as possible for any follow up to be most effective. Timely receipt of information can change the characteristics and handling needs of a claim. Identification of a claimant's insurer at FNOL instead of post payment could eliminate multiple steps & touches for the first party claims handler, reducing overall cycle time and providing superior service to the insured.

Analytic tools can match the police report data (if digitized) to the policy data, automatically raising a flag for the claims handler while simultaneously placing an indicator in the policy system to be addressed by underwriting at renewal time. Most claims systems rely on the adjuster to spot the data and send some type of notice to the underwriting department. Using analytics to collate the data and notify both claims and underwriting, increases accuracy, while eliminating steps and touches. The technology to move to this next level, this transformational approach, is available today.

This interaction of many consumers of information (actuarial, finance, underwriting), driven by claims data happens routinely in one form or another in most insurance companies today. However, the challenge to consistency and accuracy is often the human link.

While all claims organizations use data and various key metrics, either static or analytic enhanced, to manage process and resource allocation, the leap forward for many claims operations would be the addition of external data, combined with internal data and robust analytics. The key is the effective use of advanced analytics, particularly predictive modeling, to provide actionable information when using this much more robust data set. This helps claims managers more effectively link daily execution with strategy.

While the philosophy of shifting handling needs, based on new/additional information has been a core claims practice for decades, the key difference with this approach is on two levels. First, is the inclusion of near real time externally available data, which with the right vendor partner, is a one of a kind advantage. While claims has used external data for decades, it generally is procured well into the claim lifecycle (e.g. police reports, obtained in days at best, sometimes in weeks) and is usually point in time, static data, which is not always recognized or turned into actionable information.

The second is the impact of using advanced analytic tools on a regular basis (transactionally in real-time as the claim changes or perhaps nightly batch cycle) to search for, identify and weigh additional data that would provide an automated update to the handler about these changes. The broader data set, refreshed regularly, combined with analytic tools provides the claims handler the most relevant current portrait of the claim. This provides the opportunity to recast handling, potentially eliminating steps and touches, which is a win for all involved.

Claims and Exposure Triage

The impact of any of the above examples, or other scenarios that change the complexity of the claim will most likely result in a shift in claims handling needs, possibly even a shift in adjusters, based on a new skills based need. This dynamic shifting across various “swim lanes” throughout the life of the claim as new/additional information becomes available is the key to leveraging external data, now combined with internal data and powered by advanced analytics.

It will be important to also develop “intervention indicators,” that signal to a claim rep and/or manager or even to a system that the original projected path of an individual claim or exposure, has changed. Based on rules established by the carrier, alternative next steps could be recommended, including new tasks, new handling path, new reserves, etc. These indicators should also be simultaneously sent to the current adjuster and to exception based reports for managers.

Claims Impact Areas

With claims costs, indemnity and expense, comprising 65 to 75 percent of an insurers overall spend, claims efficiency and effectiveness is a key driver of overall company success. With the myriad tasks associated with each individual claim, the multiple (sometimes competing) parties involved on any given claim and the cumulative impact of improving fractionally on multiple facets in any one claim can have a massive cumulative impact on overall results. Unlike underwriting, where the insurer only has to get the information and analysis right for one policy, the insured, in claims there are often multiple parties to the loss.

This produces a complex amalgam of participants and information, which must be obtained, understood and acted upon, oftentimes concurrently. Many companies do this pretty well today. The opportunity with this new approach, is to have access to more robust data at a much earlier point in the lifecycle of the claim, allowing insurers to make more informed decisions sooner and with more consistency, which will result in claims concluding more quickly and accurately, a win for the customer and the insurer.

Challenges and Proposed Solutions

There are many challenges in effectively moving to this approach, some with people in general, some with claims in general and some company specific. Some of these challenges include the following:

Using new technology in old ways. As companies evolve and transition to new tools and new technology, it requires strong leadership and consistent follow up to assure that the new technology is used in new ways, to improve process and reduce overall cycle time and not try to fit their existing/old process into the new technology (This happens more than most will admit). As a rule of thumb, every new implementation should require that every step and touch in the current process be looked at through one of four lenses—can we improve, automate, outsource or eliminate the step or touch. The preferred outcome is the elimination of steps and touches, compressing cycle time, which improves outcomes for the customer and the insurer.

Resource allocation. It can be argued that effectively leveraging this approach can have tangible claims impacts in all four key quadrants of a balanced scorecard type strategy – the financial, process, people and customer quadrants.

Departmental silos. To effectively execute this approach requires that claims work across several non-claims areas, such as underwriting, actuarial and IT to assure the understanding, buy in and participation of these areas. This means developing the data inputs, analytics and process approaches in collaboration with these areas vs. in a vacuum in the claims vertical.

Tactical thinking. There will always be the current initiative or “crisis du-jour” that can tempt claims leaders to push such a strategic initiative off to a future date. There is no such thing as a good time to begin a transformational approach to claims process. Claims leadership simply has to manage tactical and strategic concurrently; they do it every day.

Summary

Timely, actionable information allows the claim to move forward more quickly and nimbly. This is a win for the customer and the company. Forward thinking claims and corporate executives will understand this common sense extension of underwriting approaches and practices to the entire claims handling spectrum. The data sets and advanced analytic tools to make this possible are available today. While the challenges around execution will be numerous, as is the case in the introduction of any new strategy that ripples though multiple areas of an organization, forward thinking insurers recognize these challenges and will team with the right partners to overcome the hurdles.

Special Reports /

Underwriting Claims: Data, Analytics Present Accurate Picture of an Event

Related Stories

Recommended For You