Further complicating the challenges leadership faces are continued increases in consumer expectations for products and service, an increasingly diversified market, distribution challenges, and nonstop technological advances. Bottom line, the insurance industry requires game changers to improve performance in any meaningful manner.

User-based insurance (UBI) offers one possible example of a game changer, but like all shifts it comes with it is own set of challenges. It is likely that most of the other innovations will come with their own complexities as well.

Is PAS Passé?

Legacy replacement to improve product flexibility, speed-to-market, serviceability, and quality has been an agenda item for at least 10 years now and from all indications will continue to be an agenda item far into the foreseeable future. In fact, if you combine continued rapid technological changes with shifts in markets and demographics matched by much needed innovation, the term “legacy replacement” will likely fade from the vocabulary to be replaced by a continuous cycle of modularized enhancements ad infinitum.

Cafeteria-style policies, integrated big-data analytics, wrapped external service interfaces, influential regulatory changes, distribution shifts in practices, recurring operational optimization, globalized service demands—the rate of change is not likely to slow down nor simplify, leading to the conclusion that policy admin systems will remain in flux as their new default state.

Policy administration from a technological perspective is, for the most part, a very expensive commodity function. One recent interview with an expert noted that while new systems sales per year were in the tens, there were thousands of insurance companies, implying that it would be a decade before all the companies had begun systems replacement.

In this same discussion, it was also noted the number of legacy projects that took well over five years, and sometimes ten. Reflecting on all of the aforementioned information, one has to wonder whether or not at some point the industry might recognize the exorbitant amount of human, financial, and physical capital being invested in minimally competitively advantageous processing capabilities. Even the innovations are rapidly duplicated, begging the question of whether or not they provided a market advantage long enough to recover implementation costs.

Will these fluid and complex demands combined with a talent shortage (let's not forget about the shrinking labor pool) and a limited payback period as other companies come up to speed very quickly drive the leading-edge companies to hand off the mundane aspects of policy processing to an external service provider so that the people-money-plant investments instead can be focused on competitively differentiating and financially leveraged call centers, product designs, distribution practices, agent training and support, underwriting refinements, claims fraud and pattern analysis, and other higher payback capabilities?

An objective analysis of asset utilization might support the argument that the accelerating rate of change and complexity is enough of a shift to warrant a new look at an old solution. The most agile organization might end up being the one that contracts out the huge management and resource load associated with policy admin systems implementation, enhancement, and maintenance, instead focusing their intellectual capital on market- or performance-differentiating capabilities.

Not to say that a policy admin system isn't critical, only that its functionality is marginally different across companies, making it predominantly a commodity function. Unfortunately, the perception of overwhelming uniqueness combined with parochial desires to “own” the mechanics and supporting resource of policy administration systems tend to take purchased service approaches off the table for many companies. Also, these factors often drive companies to decide to write their own systems despite the rich field of solutions.

If Not Legacy…

Memorable messaging. Variable benefit customizable products. Effective distribution. Informed risk assessment. Enhanced claims handling. Engaged service staff. Efficient operations. These are the capabilities that will drive tomorrow's leaders. Unfortunately the often decade-long legacy system projects, with their associated start and stops, vendor changes, and conversion flip-flops consume massive resources.

Addressing the above needs involves optimizing the use and understanding all available data, from what is internally stored on policy admin systems to externally acquired supplemental data. Tools and solutions have advanced to the stage where application is much less of an effort than it was a few years ago, and vendors have matured in both their service offerings and business understanding to help companies accelerate implementations with real paybacks.

Analytics' solutions have progressed beyond mining policy admin systems to the concept of “big data” encompassing tremendous amounts of information coming from multiple sources in a variety of formats. Finding small pockets of rate reductions that provide a competitive advantage with profitable customers, the uncovering of a multistate “three degrees of separation,” organized fraud, refocusing retention efforts on the most likely to stay/highest profit customers only, generating a top producer profile from a blend of internal historical data and proven seemingly unrelated external data, these all represent the positive impacts from applying analytics to data and big data.

Layering a strong analytics solution on top of policy admin data and other data sources allows carriers to focus on four areas where analytics can bring significantly amplified results. In priority order based on potential impact, they are:

Risk assessment (underwriting), including rate assignment, achieved by modeling at a more discrete level based on deeper individual and categorical risk data. ISO has introduced a tool that allows companies to move from a territory-based average rate, say at $812.50, to loss-based rating with specific policies at $1,187 and some at $438 (averages to $812.50). Homeowner rates are integrating a magnified view of the neighborhood along with supplementation by external vendors with community and additional individual information collected across a variety of modes from rewards programs to purchasing patterns.

Claims handling (claims) incorporates analytics in fraud detection, litigation management, subrogation, salvage and recovery, repair coordination as well as work distribution across staff expertise of specific claims traits, concurrent triggering of requirements, and automated escalation. Fraud referrals would come at the time of assigning claims or updating it, looking broadly at organized multi-state fraud—which is hard to catch manually—while using data to minimize claim padding and reducing false positives. This also is where fast-track claims can be identified, routed, and paid with minimal intervention, reserves set more accurately, litigation probability estimated and, if needed, reserved.

Agent effectiveness can be addressed by looking at saturation levels of agents to prospects within a discrete geography. Predictive analytics also works with prospect identification and customer retention. An increasing use of analytics looks at propensity to lapse based on historical as well as current data. Post-purchase activity is mapped for a specific client, including—where possible—integrating the performance of other similarly situated individuals. This is incremented by time-of-purchase demographic information. Mapping propensity against LEV indicates where to put the most time.

Customer awareness, which looks at the lifetime economic value (LEV) of a customer across product lines and distribution channels, while adding in spouse, child, business and ancillary line revenue to give a big picture. John Smith has homeowners, auto and liability policies along with his wife. He also has three kids with auto, one with rental, and one with specialty. One child is in medical school and one is in law school. Typical claims analysis looks at the given policies performance; LEV looks across all current policies as well as “predicts” future possible profit from the children.

Are You Listening?

Two relatively new fields of analytics that have been growing rapidly in importance and use—especially with the rapid adoption of social media—are sentiment tracking and social intelligence. In both cases, solution providers have rapidly moved to fill the need with advanced technologies targeted specifically at the social media world.

Sentiment tracking involves the development of a dashboard of indicators to show what the general “sentiment” is toward a company based on monitoring what is being said across a variety of social media platforms including Twitter, Facebook, blogs, LinkedIn, and other publicly-available information.

There are measurements of mention counts as a trend; red-yellow-green lights for whether comments are neutral, complimentary or critical; historical perspectives to provide context; and reply or escalation mechanisms to allow rapid intervention—a critically important aspect of participating in the social media world.

The social intelligence analytics are a bit controversial as they provide a means for acquiring and using non-domain, publicly-available personally-specific information for hiring, claims or underwriting purposes. The information includes atypical data, for insurance, such as shopping habits, rewards programs, magazine subscriptions, travel habits, hobbies, and anything else legitimately acquired electronically.

Individual investigations might include looking for proof of health on workers' compensation claims, validating non-smoking status, trying to find out if a person has any dangerous habits that might cause a benefit to be rated, and other similar reviews of publicly-available data. The analytics portion involves the addition of the acquired data into models being built to determine if they have any indicative value that would enhance the models' accuracy.

Companies not leveraging the sentiment tracking are ignoring a strong source of brand commentary. Whether or not a given company has a social media strategy or presence is irrelevant; they are being talked about in the social media space, and as a result should at least be listening. This one is almost a given, especially given the amplification that social media brings.

Unlike the across-the-fence or over-coffee complaints, social media gives individuals the power to talk to thousands of people about a service experience or problem. Companies not monitoring those exchanges at a sentiment level at the least are letting mini train wrecks damage their brand. The social intelligence end is more discretionary, although the number of fraudulent claims being discovered by these types of techniques is significant and profit-impacting.

Tying It All Together

Legacy policy admin system replacement is, for many companies, fast becoming more of an ongoing enhancement project than a single major push. To best realize the full potential of new or enhanced systems, and to stay on top of an increasingly complex and competitive business, wrapping the policy admin systems with a feature-rich analytics solution is necessary.

In blending the enhanced services and agility of a new or improved policy admin system with the predictive and modeling capabilities of a recent analytics solution, pay careful attention to the data. The entire process should be reviewed as each line of business is brought forward into a new processing environment.

A thorough policy admin and analytics systems implementation should review the process from end to end. What forms are used to indicate what data is needed? Have these forms been reviewed for clarity of layout and use? Have extraneous fields been reviewed and validated against current needs? What about the processes, are they streamlined, with minimal redundancy, limited handoffs, and a focus on “once and done”? Is appropriate quality checking and input review integrated into the process? More importantly, have the processes been built around the new system's capabilities versus making a new system comply with an old system's processes and constraints? Do you have a shared data dictionary that will help clear up any definitional issues? What kinds of checks and balances are in place? Are the interfaces self-checking and well documented?

Clean data has always been important; however, with analytics bringing this same data into the significantly larger roles of pricing, claims, agency, and customer, the need to ensure that the people and processes wrapping around and using the new systems for key decisions like rating, claims, agency services or customer relationships, is even more critical. While perfection is not necessary, accuracy and quality remain critical—along with a strong team with solid business knowledge. L

Steven M. Callahan, CMC, is a practice director for the Robert E. Nolan Company, a management consulting firm specializing in the insurance industry. He can be reached at [email protected].