Chicago — Attendees of Insurance Networking News’ “Insurance Business Analytics: A 2008 Reality Check” conference gathered to share insights and experiences from the front lines of the rapidly changing insurance data landscape. They also received a history lesson courtesy of Matt Josefowicz, director of the insurance practice at New York-based
Josefowicz noted that there is abundant evidence of a thriving insurance industry in ancient ruins near Rome, and that despite a couple millennia of technological advancement, the mission of insurers haven’t changed all that much.
“When all the discussion about business analytics started a few years ago, my question was ‘isn’t that what we’ve always been doing for the 2,000 years—looking at data, analyzing it and trying to make predictions?’” Josefowicz asked.
Yet, Josefowicz said the advent of business analytics, a rubric including both business intelligence and predictive analytics, was indeed epochal, shifting the insurance industry from an age of data scarcity to data abundance.
“The changes recently in the amount of data, the sophistication of tools and our ability to model it really have resulted in a qualitative change that deserves a new name,” he said. “The amount and quality of external data available to insurers is skyrocketing, but the good news is the analytic power available to insurers is keeping up.”
Accordingly Josefowicz sees insurers move from a tradition of gathering data to spending more time validating and analyzing it.
To make this transition, insurers need to make wise investments in the correct tools.
“There’s a huge difference between the investments by organizations that want to be competitive, and those just want basic, functional capability,” noted Mark Gorman, principal, St. Paul-based Mark B. Gorman & Associates LLC.
But in analytics, as in life, money can not assure contentment. Among the many issues that can stymie an analytics implementation is the “swiss army knife” conundrum, where the tools have become so advanced carriers don’t fully utilize them.
Yet, getting the proper tools in place and knowing how to use them is merely part of the challenge. Insurers must strive to mesh these tools with both data warehouses and production systems, which provide the feedstock for analytics.
Josefowicz said that in a survey of carriers recently conducted by Novarica, only half of respondents expressed satisfaction in existing capabilities to support business inteligence. “The fact that only half of carriers surveyed felt confident in the integrity of data coming from their production systems is a little frightening,” he said.
Moreover, once installed, an analytics systems needs to be wedded to a data model and, perhaps more importantly, it needs the buy in of business users.
“You can have the best data infrastructure in the world, but if you don’t have a clear sense of how users are going to use that data to run things more effectively, it’s all for naught,” Josefowicz said. “There’s no use in delivering somebody insights if there’s no way to act on it.”
Gorman said data, although long the bane of many on the business side, needs to become part of the cultural DNA of an insurer for an analytic investment to bear fruit.
“The cultural change needed to do predictive analytics requires a visceral understanding of the data by the business,” he said.
However, many in attendance acknowledged that this understanding does not come easily. Often, it faces a bulwark of opposition in the form of entrenched business practices. The industry has long elevated operational excelence over thought leadership, Gorman opined. Many insurance companies have been stable for so long t hat the impetus to operate differently has been minimal.
“The culture of using enterprise data to change operations is not a strong culture in insurance,” Josefowicz said.
Indeed, Gorman recalled a conversion with a claims adjuster who deemed analytics “all well and good,” but insisted it was superfluous because she could detect fraud by merely examining handwriting.
Joel Appelbaum, CAO of
To remedy this, Gorman said an analytics initiative needs a champion within the organization with sufficient clout to get the program off the ground. This entails selling the initiative to upper management, which is no easy sale because the tangible benefits for BA solutions don’t have a tendency to show up quickly. Thus, it’s imperative that a demonstrable link exists between the technology and a future operational improvement.
“The business case has to be driven by what is going to be the change in business behavior,” Josefowicz said “Guess who is going to get blamed for an expensive project with no ROI?”
Gorman said CIOs have to come to view themselves as data guardians, not stewards.
However, Gorman stressed that a strictly top-down approach from the C-level is not advisable. Rather, he advocated a middle-out approach, where business units take the lead, much like how fast-food franchises are supported by a centralized hub that handles R&D, standards, quality insurance, personnel and algorithms.
Such is the set up at the Chicago-based
Andrea Marks, executive director, health intelligence, heads up the analytics efforts for the association, which is comprised of 39 separate plans.
Marks said it’s not as difficult finding people with the technical expertise for analytics as one might expect. “The challenge is more developing an analytic R&D team—the people thinking through the predictive models,” she said. “The approach we’ve taken is not necessarily looking for people with a background in healthcare, the business aspects can be learned. What we do look for are people who can look at data differently.”
To round out her team, Marks said she relies on people with internal knowledge among the organization's 39 plans as well as vendors for their expertise.
Despite these technical and cultural challenges in front of an analytic initiative, Gorman, who has conducted independent research on analytic use by insurers, said the use of analytics is broadening across the enterprise, not just in traditional strongholds like claims and underwriting.
Yet, Josefowicz acknowledged there are limits. “When you get into marketing and customer service, the ability of insurers to leverage analytic capabilities to create business value is much more limited,” he said.
Gorman concluded that the overall trend is toward greater use of analytics and not just at large insurers. “It’s widespread and underneath the radar with small and mid-market carriers,” he said. “There are many organizations doing pieces of this, it’s much broader than you know or believe.”
Exclusive content only available on InsuranceNetworking.com
