Predictive Modeling: Strategic Opportunity for Agents

Predictive analytics is fast becoming an essential tool for insurance providers. Carriers are using predictive analytics for segmentation, underwriting, marketing and even claim operations.

Here’s what this means for agents: You can grow by leveraging carrier-based analytics for cross-sell and up-sell campaigns; you can attract new business that fits carrier appetites; and you can manage customer portfolios more effectively through customization.

Carriers and agents already have witnessed a trend to introduce models for marketing and rating tiers. This trickle of models is likely to become a flood of sophisticated and highly actionable information. To thrive in this environment, insurers and agents must better understand the driving forces behind a momentous shift in insurance marketing, service and operational management.

Property-casualty predictive modeling has grown steadily for several years. Initially, a few personal lines direct writers built entire business strategies around predictive rating models. These early-adopters succeeded in “skimming the cream,” leaving less attractive risks to traditional competitors.

Mainstream carriers took notice. With dramatic increases in both information resources and computing power, they are busily hiring and training statisticians, mathematicians and actuaries to bolster analytic expertise. Or they are contracting with third-party providers to build predictive models—and not just for underwriting and pricing. Models today are being deployed for marketing, fraud detection, claims management, subrogation and audit control.

How is predictive modeling defined? Simply put, it is both the science and art associated with exploiting statistical analysis of past events to predict future outcomes. This sounds remarkably similar to what actuaries have always done: look at risk characteristics to predict loss levels for similar future risks. However, recent capabilities to store and manage vast amounts of data—and innovative analytic techniques to evaluate that data—enable these predictions to be much more precise and powerful. For example, traditional personal auto rating territories might have been derived from a combination of several counties or, at best, several ZIP codes analyzed together. Modern predictive techniques can differentiate risk levels in areas just a few city blocks apart.

Significantly for agents, insurers are realizing they can apply these same techniques to marketing, retention and service management. With help from agents, companies are re-examining their books of business. One carrier may target the mature, wealthy segment of the baby boomer market; another might target the young, recently independent customer. Companies can make special efforts to cost-effectively attract a prospect category or relinquish it quickly.

The success of traditional target marketing programs often depended on how many prospects responded to outreach efforts. Today, insurers and agents must expand the analytic dimensions of marketing programs by evaluating precisely how many of the right customers respond and why. Locating customers that optimally fit products, services and price points is much more effective now through the power of predictive modeling. Ultimately, such prospects become happier customers, benefiting from tailored products and point-of-service intimacy. Moreover, insurers and their agency partners can use predictive capabilities to attract and keep more loyal, enduring customers. These initiatives are driven by the smart deployment of predictive analytics—impacting insurers’ interactions with their agency forces.

Clearly, predictive technology influences the range of expectations and opportunities facing agents. Insurers are increasingly relying on their agent channels to successfully implement the new focus on customers’ personalized needs. Companies also are relying on the support and partnership of their agency force to effectively realize the promise of their analytic assets. In most cases, agents are better positioned than insurance company personnel to explain these new market transformations—and, more importantly, the resulting value—to customers. Insurers looking for a competitive advantage are investing in predictive modeling to create greater efficiency in their operations, to enhance their products and services and to develop more coherent service and retention strategies. In fact, possessing a strong capability in predictive modeling and deriving real value from the multidimensional analysis of data are quickly becoming competitive imperatives.

The race for effective predictive modeling will ultimately result in cost-effective products and new services for insureds. Insurance companies that understand and adopt these trends are reshaping the industry. But insurers cannot implement these new strategies on their own; they need agents who understand predictive analytic tools and can execute customer-focused strategies. By embracing the power of predictive analytics, agents will be the key conduit for product value and exemplary services between preferred insurance markets and satisfied policyholders.

Phillip Hatfield, a licensed agent, is vice president of product development for ISO Innovative Analytics in San Francisco.