are the basis for optimal results for the adjuster, carrier, and
client. The combination of digital photography on the adjuster
end and OCR on the carrier end turn visual images into data
that can be handled like text, fed into predictive models, and
processed by the systems.
Once in house, a combination of business process management software, core claims processing system, and analtyics take
over to optimize the handling of the claim while also minimizing
risk and analyzing for fraud. Starting at the front of the process,
the system can identify, segment, prioritize, and score the claim.
Data mining is used to cluster and group claims by loss characteristics (such as loss type, location, time of loss, etc.).
Claim and insured party characteristics are then used to
score, prioritize and assign the claim, improving the consistency of claims routing and ranking while improving the
quality and speed of the claim through the system. Studies
show that claims cost can increase by as much as 40 percent
if FNOL is delayed by four days. The analytics also assesses
for probability of litigation, given estimates indicate litigation
costs two times over normal. By identifying claims that are
likely to result in a suit early in the process, they can be routed
to more experienced adjusters.
Other front-loaded areas where analytics can play an
important role include workload planning, cashflow patterns,
claims duration, and loss reserve forecasts. Throughout the process, as new information is received and the claim matures, the
process of determining litigation propensity and fraud scoring
are repeated with escalations automatically done when a change
in status is noticed. During the process analytics are often used
to identify salvage and subrogation opportunities, indicate
deviations in prior results, and provide best case scenarios for
the final steps in the claim.
The largest opportunity for analytics rests with fraud
identification. Given the increasing growth in multi-state fraud
rings, the cost to the industry continues to increase well into
the billions representing far more than the occasional reference
to “cost of doing business.” Most carriers have become adept at
utilizing a rules library to capture fraud based upon known patterns of previous frauds or individual and aggregate anomaly’s
that appear in the data.
Some of the more advanced carriers are using knowledge
discovery, data mining, and predictive assessments to detect
complex patterns hidden in the details of large volumes of claims.
These are often the fraud rings. A rapidly growing field is in the
area of social network analytics used to determine associative
links not otherwise discernible, resulting in relationship or entity
maps that can be reviewed with experienced investigators.
These are all enhanced by the advances in text mining where
analytics systems are trained to search images, faxes, and other
handwritten or typed documents for key words leading to important claims information. The findings are then routed to the
adjuster for review and determination of value.
For subrogation, analytics offers a number of advantages
throughout the process. Typically, given a lower priority on
the workload, the task ends with a good bit of rekeying, lack of
good details on historic information, possible cost containment
options, and tracking of results over time across vendors and
outside parties. Recapturing some portion of the estimated 15
percent of claims that miss subrogation opportunities can have
a meaningful and measurable impact on profitability.
Determining additional liability, found in a pilot test to be
possible in approximately 60 percent of the claims reviewed,
could result in a material reduction in payouts. Recovery and sal-
vage analytics combine access to a wide range of databases, access
to prior history on similar negotiations, and the ability to crunch
large volumes of resulting data to determine the optimal solution.
Lastly, for the claims that end up in litigation, analytics provides
the basis for determining precedence, frequency, standard practice
at the least as well as doing deep dives into publically available data-
bases, postings, social media sites, and internal systems.
A recent survey of senior level claims executives indicated
that 85 percent felt their existing claims systems lacked the flexibility and functionality necessary to sustain growth and deliver
value, with 78 percent of the respondents indicated an upgrade of claims systems in process. The underlying issues were
validated by the 32 percent that stated claims relies on more
than five different systems with the core system of 54 percent of
respondents being over five years old.
Claims systems replacement has finally reached the top tier
of many companies’ priority list. Unfortunately, as noted in
many different studies, fewer than 20 percent of replacement
system projects actually succeed as planned, while approximately 50 percent are challenged by significant budget and schedule
overruns and the remaining 30 percent end up considered failures. Those are not very appealing odds given the unforgiving
market and the short window of time.
That said, when leaders work as a cross-functional team
with a single, clearly stated and shared goal using the principles
of transparency, objectivity, focus, and discipline, success is
achievable. Enough experience has been gained through the
years to change that 20 percent success rate to 80 percent or
higher by simply leveraging the existing body of knowledge
on what works, what to watch out for, how to govern, and how
to maintain a real-time line of sight on progress and risks. For
those facing this challenge, the answers are out there; take a
moment to look, ask, and reflect. It could make the difference
between being in the 20 percent or the rest. ITA
(Steven M. Callahan is a practice director at the The
Nolan Company, a management consulting firm
specializing in the insurance industry. For 40 years,
The Nolan Company has helped insurance companies
achieve measurable improvements in service, quality,
productivity and costs through process innovation and
effective use of technology. Callahan can be reached at