Insurance Industry Analytics: Where Are We Headed?

By Scott G. Stephenson

What direction is the insurance industry taking when it comes to analytics? The answer might depend on the nature of each line of business.

From many published accounts — whether by Verisk or other industry experts — it appears insurers of all sizes are adding or planning to add analytics expertise to their business and technology arsenals. Personal auto seems to have the lead in analytics adoption, but homeowners has picked up momentum too. Carriers might yet see potential for modeling in commercial auto, but with commercial and nonstandard products — less homogenous lines — manual underwriting is not likely to go away.

Insurers realize analytics capability can be both a competitive opportunity and a challenge. While large carriers with more resources may have an advantage in analytics adoption, they also face the difficulties of analytics implementation on a larger scale: the bigger the project, the greater the risk. At the other end of the spectrum, small and mid-tier companies are coming up to speed through agile management.

The analytics process is evolutionary: Stakeholders must be aware that implementation isn't a singular project; rather, it's an iterative initiative. And the better the feedback, the better the results.

Carriers of all sizes are developing their own analytics capabilities or achieving jump starts by working with solution providers to create the needed infrastructure for sophisticated analytics programs. While some insurers may believe the cost of external analytics support is still prohibitive, others have found that outside providers actually offer economies of scale and time-tested industry expertise, indeed leveling the proverbial playing field. Realistically weighing the analytics investment — whether for in-house or vendor options or even a combination — insurers need to factor in the cost of lost opportunities in underwriting, actuarial, fraud prevention, marketing, meeting customer expectations, and simply staying competitive in an increasingly complex business.

When carriers turn to solution providers, the reasons are as varied as the carriers themselves; and analytics expertise, rather than carrier size, is often the determining factor. Insurers may need guidance to manage the complexities of an analytics project, prefer to buy rather than build a system, or require augmentation of internal human or data resources. While cost is always a concern, access to a packaged solution that integrates smoothly and swiftly into a carrier's environment can, at the least, match the competition or, at best, provide competitive advantage.

It may well be true that certain books of business in specific lines don't warrant the investment in analytics, but that assessment depends on a whole range of factors, both objective and subjective. Not surprisingly, each carrier seems to exhibit its own appetite for analytical capabilities, as do specific lines of business within each insurer.

Generally speaking, carriers have come to realize the value of analytics to support underwriting and rating. Traditional underwriting procedures can and do coexist with sophisticated modeling algorithms by testing and reinforcing each other, ultimately yielding better underwriting results. As a decision-making or supplemental tool, a combination of the two approaches enables greater efficacy of segmentation in risk characteristics and pricing.

Beyond underwriting, the benefits of analytics for all aspects of insurance companies are well documented: increased sales, decreased expenses, higher margins, added speed, better fraud prevention, improved claims settlement, enhanced carrier-agent relationships, smarter enterprise risk management, more effective catastrophe management. And that covers only a fraction of the value that analytics offers.

Interestingly, two companies with the identical model will often get different operational results. Nevertheless, though the end result may be specific to one company, best practices are applicable to every company.

The following are tried-and-true strategies for analytics success:

  • Many people resist a shift toward analytics, especially when they've been comfortable with conventional methods. To move the company forward, educate all stakeholders about the role and capabilities of analytics.
  • Be truthful and frank in assessing analytic capabilities, understand the timelines of varying kinds of implementations, plan for a viable budget (high-six to low-seven figures is common) — and realize the value-add will be worth the effort.
  • The analytics process is evolutionary: Stakeholders must be aware that implementation isn't a singular project; rather, it's an iterative initiative. And the better the feedback, the better the results.
  • Partner predictive modelers with business users. The specialists on both sides must work together to ensure the outcome meets the requirements of the business.
  • Consider the organizational structure and any changes needed to bring analytics closer to the business units or to develop a centralized shared service environment.
  • While analytics development often rises from the bottom up — whether underwriting, actuarial, products, marketing, or claims — initiatives benefit from top-down support involving multiple departments and faster implementation.
  • Executives, product managers, underwriters, actuaries, and IT personnel are just a few of the stakeholders who need to be on board with an analytics initiative — perhaps especially because the analytics labor market is highly competitive today and personnel with analytics experience may be scarce.

Innovation produces more effective analytics, and innovative analytics produces a more effective insurance company.

Innovation produces more effective analytics, and innovative analytics produces a more effective insurance company. In personal auto, differences in state regulations and an expanding data pool create the need for ongoing refinements and augmentation. Enhanced geographic data is energizing the use of analytics in the homeowners arena. More geographic data means greater segmentation and improved risk identification. Other technologies such as telematics, while seemingly cost-prohibitive for some carriers, are gradually gaining ground, too.

When assessing the state of analytics and the insurance industry, the analytics train has already left the station. Insurers of all sizes are laying data and analytics tracks to facilitate their business objectives — with measurable results. Furthermore, just as passengers appreciate a more efficient transit system, so do policyholders benefit from a more efficient insurer that offers better pricing, products, and services. In return, the carrier gains by satisfying its customers. We're now seeing the momentum for analytics build — taking the industry to a more innovative, competitive, and successful destination at the end of the line.

Scott G. Stephenson is president and chief executive officer of Verisk Analytics.