The Reality Versus the Dream of Insurance Carriers

Insurance is difficult. Many years ago, Scott Peck quoted in his book “The Road Less Traveled” that “life is difficult”. He went on to explain that once people accept that fact, things become a whole lot easier. While simplicity is achievable for people, not so much for insurance.

At its core, insurance is a simple concept. However, just as effective parents learn to separate the child from the behavior, it is critical to recognize that “insurance” needs to be separated from the “administration of insurance”.

Insurance is simple: issue a written document that lays out terms of coverage, premium, exclusions, and conditions and then pay claims accordingly. Right. And, if you knew a risk perfectly, the perfect system could be created to track the policies issued, along with the corresponding terms and conditions. Likewise, you could track claim activity and pay claims just exactly as the policies had been priced. Wrong.

The Effect of Judgment

Administration of insurance becomes very difficult because of one word: judgment. When terms and conditions are bound with an insured, the agent, insured, and underwriter each use their own judgment in assessing if the coverage and premium is proper for the underlying risks and exposures.

Likewise, when a claim occurs, an adjuster applies judgment during the process of adjusting and settling the claim. If the claim involves bodily injury or litigation, multiple layers of judgment have just been injected into the process through the multitude of medical and/or legal professionals that will cast judgment about injuries, treatments, coverage(s), and causes.

Insurance systems are great at providing a means to capture information about policies and claims that is unrelated to judgment. For example, names, dates, addresses, type of vehicle, and so forth. That type of information is often referred to as metadata.

Additionally, over the years, the insurance industry has adapted to the need for judgment to be documented by creating codes that can represent events and activities. For example, a worker trips on a rug slips on a piece of paper and then falls down. A claims representative will use judgment to assign a code to this event as a “Trip”, “Slip”, “Fall” or “Slip and Fall”.

The problem is that each of those codes is right, to some degree, and is also wrong, to some degree. Why? Because, importantly, history has shown that claims representatives are more often than not inconsistent in the way they apply their judgment when adjudicating a claim and describing information about the claim.

This doesn’t make them bad or wrong. It is just the simple reality that people apply judgment in varying ways. And, their judgment can be influenced by a litany of factors.

The application of those types of codes creates what some people refer to as semi-structured data. The codes enable systems to capture at least the essence of critical events and activities happening with an insured’s account, a claim.

Go back to the basic premise: if you knew a risk perfectly, then the perfect system could be created. Of course, it is impossible to know the risk perfectly. Not only is risk abstract, nebulous, and difficult to quantify, it is constantly changing. The risk you thought you understood yesterday – around which you designed your system – is not the risk that you have today nor the one you will encounter tomorrow.

The dream of insurance carriers – a would-be utopia – is that all systems work together to seamlessly provide every piece of information needed to make great decisions. The reality is that because of the prevalence of judgment in the insurance mechanism, systems can provide only the basic information needed for policy and claim administration.

Before we turn to hope, let me put the final nail in the coffin of despair. In the near 34 years I’ve been working in the insurance industry, I have seen company after company get caught in this cycle of trying to migrate to a new system that would better handle the nuances of the risk being encountered in the present. The reality is that systems simply aren’t equipped to handle judgment.

The Hope of Today

So, let’s go back to judgment. Systems do not produce results. People do. And, the people with the best judgment will produce the best results.

In the normal flow of insurance, there are people using judgment from start to finish. Consider:

• Insured’s and agents provide descriptive and numerical information to support the risk for which they want coverage.
• Underwriters review information, they research and they provide a quote for insurance coverage. As part of that process, they document their decisions and supporting rationale (i.e. their judgment) based on the research they’ve completed.
• When a claim happens, the adjuster documents the decisions and supporting rationale (i.e. the adjuster’s judgment) for the claim-related needs and decisions that are communicated to the claimant.
• Furthermore, adjusters document decisions and supporting rationale for medical diagnostics, case management, litigation support, and a variety of other factors relevant to adjusting a claim.

All of the documented information of the various constituents can be thought of as sentence data. Sentence data is unstructured data that exists along with a lot of other content that has no structure, such as pictures, video, and voice. It is original content documentation of what has happened, is happening and why, who’s doing what and when, and a lot of other critical information about the core activities of within or related to the insurance process; whether underwriting- or claim-related.

Sentence data is the unequivocal, absolute best source of information about risk as it is happening right now, the associated insurance coverage that may or may not apply and all of the related activities associated with underwriting and claim related processes.

Encapsulating and understanding that unstructured data provides key insight and a more robust understanding of critical elements to assess the full circle of the insurance process and better assess and manage risk and exposure, and leverage that understanding to drive better results. The hope of today lies in the fact that it is now possible for all of the unstructured sentence data to be converted into measurements and metrics of critical operational Events and Activities that drive margins and improve customer satisfaction.

Operational Events and Activities

When a person manually reads documentation, they are looking for Events and Activities that happened, did not happen or may happen. And, based on a sequence of Events on a claim, for example, the reader is looking for the adjuster to have completed a set of Events or Activities.

Using technology to identify critical Events and Activities changes everything about measuring and understanding risk and the people underwriting policies and paying claims.

A practice that has existed for as long as insurance is auditing files – whether they are underwriting files or claim files. Some companies today have hundreds of people in audit or quality assurance functions that spend their time reading files. Not only is reading files ineffective and cumbersome, but sampling files is also way too inadequate for the purpose of measuring an underwriter or adjuster’s judgment.

Seriously, if somebody is handling 150 or more files, does looking at 3-5 files really provide enough evidence to substantiate their performance? By using technology, every note they wrote can be considered in measuring their judgment on the critical Events and Activities that will influence the successful outcome for a claim.

Using measurements and metrics created from sentence data does not in any way mean that the days of file audits are over. Rather, by measuring critical Events and Activities for the entire population of records, it becomes possible to identify outlier files and people performing above or below a targeted performance level.

With this new information about Events and Activities, poor judgment can be identified in a matter of days or weeks compared to months or quarters. Likewise, problems or opportunities with agents and insureds can be spotted quickly and solutions created long before competitors are aware of the opportunity.

Measure to Improve

It is a long-standing axiom that what can’t be measured, can’t be improved. Arguably, the inability to measure the judgment of people underwriting policies and adjudicating claims is the primary reason that the insurance industry has been unable to produce breakthrough results.

While measuring somebody’s judgment is theoretically impossible, measuring their performance based on the Events and Activities in the sentence data they wrote creates a level of understanding of their judgment and a measurement thereof that has never before existed.

Insurance is difficult and insurance carriers need effective systems. But systems don’t produce results, people do. And, people with the right judgment produce good results. The insurance carriers that adapt to using their sentence data to assess and measure results based on the judgment will find breakthrough results long before the competition. Those results will drive improvement in the judicial processes of their people and also provide an opportunity to improve end-to-end insurance margins.