It's nearly every day that I'm reminded what a great industry insurance and risk management is. Technology has brought us both new types of risks and new ways to mitigate or understand them. People continually find ways to bring new ideas to reduce risk, ideas that are spawned by creativity, and often nowadays by analytics.
There are new things happening all the time, from inspired writings on LinkedIn, industry newsletters, and newly authored books to inventive startups seeking to turn our insurance space inside out and to students graduating with insurance degrees or being confirmed as a Chartered Property Casualty Underwriter or other industry designation.
To be perfectly honest, I have what my kids call "FOMO" for "fear of missing out." Therefore, I have trouble NOT paying attention to new thoughts, new opportunities, or new ways to approach old problems. And let's not forget new problems. That's the attraction of analytics for me. My IRMI article from a year ago, "Insurance Company Data-Driven Opportunities," originated in my thoughts to stimulate the readers' creativity about leveraging the latest and greatest analytic technologies. For me, it was a strain coming up with all the ones I listed. I had seen many solutions in action and made notes about others from conversations with colleagues and customers. But others (transparency here) came to me out of the blue as I was writing.
Any Data Is No Better Than No Data
Who ever thought about managing claim defense costs using analytics? As can be imagined, this is not a new concept, as claim defense costs are a very large insurer expense. Because insurers want to manage this expense, there are several "e-billing" software providers that do help insurers reduce their defense costs. Insurers use these software tools to make sure they're not getting double billed or being billed at the wrong rate or for other tasks, such as performing basic "auditing" checks. I am sure I am missing some of the other "discoveries" these systems provide to reduce costs, but they are not analytics in the true sense of the term.
Analytics require clean data and meaningful data. E-billing systems rely on the human coding of law firm efforts to meet the parameters of Uniform Task Based Management System (UTBMS) codes. If an insurer has 100 law firms it works with, it could have 150-plus people doing coding at these firms. It doesn't seem there would be consistency from person to person or firm to firm since the coding is a manual and subjective process. Neither application of code is right or wrong, just different. Most data scientists or data governance experts would challenge these codes as recorded and as not providing clean data, although they could be very meaningful if they were.
The data in a law firm's system (and carried over to e-billing) just may be incorrect. For example, one company exploring UTBMS codes for use in analytics would find appeal-related costs (L500 series) sometimes coded before the trial even takes place. Part of the problem is UTBMS codes may not be applicable for each case or else are not what they're designed for.
Lastly, the UTBMS code set is simply not granular enough to support analytics. The codes were never designed to link expense records with fee records. Take, for example, a deposition. Travel costs that are just shown as part of a case do not reflect how many depositions there were or which expense is tied to which deposition. With no denominator, it's impossible to know if the expense provided value to the outcome of the case. While these codes have served a purpose, there's a fundamental flaw in using them for analytics. There is also no outcome tied to whether the depositions (type or number) affected the outcome of the case.
Inquiring Minds Want to Know
Insurance claim professionals want to better understand where the claim dollars go. Insurers are being squeezed for every dime of the premium dollar, and claim costs are the biggest expense. Reference this 2015 survey of 94 property and casualty insurers about managing defense costs. The list below has the top metrics measured or deemed to be important by the insurers' surveys (by percentage of respondents).
- Average total cost of case (loss + legal costs): 84%
- Legal expense per case: 82%
- Cycle time (days to resolution): 82%
- Loss per litigated case: 77%
- Allocated loss adjustment expense (ALAE) as % loss ratio: 71%
- Allocated legal loss adjustment expense (ALLAE) as a % of loss ratio: 63%
- Average or median bill rate by claim type: 62%
- Staff vs. outside counsel outcomes: 51%
- AFA outcomes by fee type: 44%
Source: Denise Johnson, "The 3 Most Important Metrics for Measuring Performance and Managing Litigated Claims Costs," Claims Journal, December 10, 2015.
These are good questions and, if answered effectively, will lead to claim defense cost savings. However, there is an entire world of unknowns out there because few have enough data available or in the right form to think in such detail. If we had the right data available, wouldn't it be wonderful to answer the following questions?
- Is it an effective strategy to file a motion for summary judgment (MSJ) in a particular venue or with a particular judge, given our historical success rate? How much does it cost to file an MSJ?
- What is the average cost of an expert deposition, and are we taking more of them now, or has the average cost per deposition increased, or both?
- What is the optimal lag between preparing our defendant for his/her deposition and the deposition itself, or any?
- Do we tend to get a better outcome when the lead attorney's hours represent at least X-percent of the total hours spent on the case?
- How much does it cost to have our defense firms comply with our 90-day claim summary report, and does the compliance rate correlate with outcome of the claim?
- Can we develop a more cost-effective strategy for our record retrieval on court reporting costs?
Source: Chad C. Karls, "Big Data Analytics—A Practical Application for MPL Insurers," Inside Medical Liability, 2015 Fourth Quarter, pp. 24–25.
Of course, there are many more answers possible if the data can be extracted from the claim file. Claim files are rich with information, but it's just hard to get at. Medical professional liability is clearly an opportunity for an analytics application, but it's only one line of business. Consider most of the long-tail lines or situations where multiple insurers are potentially involved, such as construction defect or product liability. I have consistently stated that granular data available for analysis ultimately will provide a competitive advantage for those insurers willing to capture and create this data. While creating this data may be a challenge, the good news is the insurers are already collecting the information needed to create this treasure trove of data in the form of their defense attorney invoices. By successfully leveraging this data, insurers will be able to better understand and thus manage this growing expense.
Conclusion
The bottom line is to keep an open and imaginative mind when it comes to analytics. There are new techniques becoming available all the time as well as improvements in software making the tried and true techniques much more powerful.
In the above example, the application of business directed and experience text mining have already proven to go a long way for those willing to make the leap. The return on investment can be as strong as or stronger than any other application of analytics for an insurer. Claim expense savings go directly to the bottom line and in a hurry!
Opinions expressed in Expert Commentary articles are those of the author and are not necessarily held by the author's employer or IRMI. Expert Commentary articles and other IRMI Online content do not purport to provide legal, accounting, or other professional advice or opinion. If such advice is needed, consult with your attorney, accountant, or other qualified adviser.
