"Many companies have hired good analytics people but haven't 'productionized' their output, instead creating a kind of legacy production process where they repeatedly go back to the table to tackle some ad hoc problem," Lucker continues. "What you need is to have the foundational data being routinely produced inside the business; you have to have a scoring or analytics engine that generates the analytic results, and those results flow to a business rules engine that translates into actions that can be executed throughout the technology infrastructure."
Few companies have succeeded in bridging that gap for more than point solutions, according to Lucker. Most carriers have applied analytics to nearly every aspect of the claims cycle, he says, but now they are reevaluating what can be realized in areas that had been difficult to automate because of data hygiene and other problems. Carriers are now casting the net wider for higher ROI solutions, Lucker relates. "The key to these solutions is that they should fit within a more holistic strategy for end-to-end claims lifecycle management, with analytics at the core," he advises.
Guiding Resource Allocation
Narragansett Bay Insurance (NBIC), a Pawtucket, R.I.-based homeowners insurer with $62 million in annual written premium, is paving the way for such a strategy. According to Jose Trasancos, NBIC's SVP of research and product development, the carrier's plan is to weave analytics into the fabric of the claims processing environment in order to be able to structure the use of resources applied to claims.
"We should be able to understand the expected outcome of every action and the competence intervals around that expected outcome," he says. "The ideal is to bring objective data to bear at every decision-making node in the claims process."
Trasancos sees the discipline of analytics in insurance as moving away from a one-size-fits-all approach toward managing claims severity with analytics. To illustrate, he refers to one of the early metrics of claims analytics: inspecting a damaged vehicle within nine hours of an accident. While taken broadly, it was a useful metric, Trasancos acknowledges. However, only a subset of all cases will benefit from early intervention.
"If you get first notice of loss at 11:00 p.m. and the car is in the driveway and drivable, sending an adjuster within nine hours is a fool's errand -- you're chasing a number at that point," he says. "If a tree falls and takes out a section of fence, you don't need to be there -- if it hits the house, yes."
NBIC is aiming at using analytics to judiciously manage resources to ensure the best outcome. Analytics can help decide when it makes sense to apply costly resources, such as dispatching a cause-and-origin engineer within hours to a claim site, or when to forbear from doing so when no efficiency or customer value is likely to ensue.
"You can triage claims a number of ways, and at different points in the resolution process, with a view to continuous improvement," Trasancos says. "You evaluate a hypothesis, and if it looks promising, you respond; if it bears no fruit, you look elsewhere."
Trasancos relates that NBIC has adopted tools from Cary, N.C.-based SAS and the SAS/STAT statistical library for analytics. The carrier also uses JMP (Cary, N.C.) data visualization tools. And this past fall, NBIC added Honolulu-based BlueWave Technology's PipelineClaims system. "Operationally the BlueWave system is very intuitive and, in the event of a catastrophe, easily deployed to a large external network of service providers," Trasancos shares. "It is also an excellent foundation for the claims data environment. The operational flexibility of the system will help us bring analytics to bear in the claims process."
Recognizing the importance of the data foundation of analytic processes, NBIC has undertaken what Trasancos characterizes as a focus on data discipline in order to extract the greatest possible value from its technology investments in analytics. "To the extent that we can make data easier to consume for analysis, we will accelerate the process of weaving analytics into the fabric of the claims process," he says.
More Data From the Start
Trasancos reports that claim analytics at NBIC begin with the acquisition of a piece of business. "We inspect each property, inside and out, and have a structured way of collecting information on every risk as it relates to mechanical, electrical, plumbing, etc.," he relates. "We also make qualitative observations and keep a photographic record."
As a result of collecting more precise data from the inception of a policy, NBIC is able to make more accurate reports from first notice of loss and support a more granular analysis and allocation of resources, Trasancos asserts. "This is a valuable source of information not only for the claims organization but also on the risk side," he says. "We have first-hand knowledge of the characteristics of the risk, and that has proven quite useful in terms of determining the kind of remediation, reconstruction and repair that we can bring to bear."
Charged with driving analytics to the next level within his company's claims organization, Farmers' Vandenberg emphasizes the cultural dimension of claims analytics, in addition to building qualified analytics staff and effectively embedding predictive models in production environments. "In the claims world we're not keen on completely automating things," he comments. "Fundamentally we are an organization of individual claims adjusters talking to claimants and making judgments along the way."
Vandenberg insists that optimal use of analytics in claims requires embedding analysis as a way of providing support for the decisions that adjusters make. He stresses that any claims strategy is only as good as the company's ability to manage these small decisions in a vast sea of decision-making.
In addition, carriers must not make the mistake of looking at both adjusters and end customers as mere subjects of the analytical process, according to Vandenberg. "Don't think that customers are not participants in your analytic strategy," he counsels.