The mere mention of a data warehouse has often caused even the most unflappable insurance carrier executive to break out into a cold sweat.In recent years, the failure rate of data warehousing projects in the insurance industry has been a dubious distinction, owed in no small part to insurers' inability to effectively plan and execute such projects. Considered "data rich yet information poor," insurers have struggled to create data warehouses that can truly tap into the power of their customer data.

But an emerging initiative known as business intelligence-on which data warehousing serves as a cornerstone-may be helping insurers put these responsibilities in better perspective. Business intelligence has become a rallying point for insurers to rededicate themselves to best-practices data warehousing, data mining and data cleansing, strategies marked by clearer visions and higher hopes.

Business intelligence solutions cast a wide net over an insurance operation, enabling users to leverage data to reduce internal costs as well as generate revenue. Investments in this space span the spectrum of what insurers deem as priorities, such as claims management, regulatory compliance, underwriting, marketing, fraud detection and litigation.

However, saddled by their lack of skill in collecting and analyzing data, insurers haven't fully tapped the power of business intelligence.

Nevertheless, industry experts say the environment is improving within a dual-track dynamic of better strategizing by insurers and more sophisticated technology. Furthermore, experts say insurers recognize business intelligence as being core to their operations and are responding in kind.

"Investments in business intelligence are definitely top of mind with senior executives," notes Matthew Josefowicz, senior analyst for Boston-based e-business consulting firm Celent Communi-cations Inc.

"The chief factor driving business intelligence investments is that insurers realize the need to get a mastery of their data. When investment income was pouring in, insurers could cover up for a multitude of sins, including their inability to efficiently collect and analyze data. Now that investment income is scarce, there has to be a greater amount of accountability with data collection and analysis," he adds.

Tom Chesbrough, founder and executive vice president of Frisco, Texas-based Thazar, a Skywire Software company, sums up the matter this way: "There are billions of pieces of data flowing from insurers' systems every month. To this point, the approach has been to manually sift through mounds of paper to identify the areas that require the most attention. Now, automated front-end technology customized and modified for specific users is driving the effort."

Bigger isn't better

Large insurance companies have long been saddled with a dilemma: the bigger they get, the more data that must be reconciled. "Insurance is about using data intelligently, but insurers have not always executed this well. To compound matters, insurers have always been saddled with an overwhelming amount of data," Josefowicz says.

The acquisition pace of several years ago only exacerbated an already challenging problem for insurers. Insurers had an appetite to acquire not only other insurers but also a variety of managing general agencies (MGAs) that, in essence, are small-size insurance companies based on their assets and resources.

In 1999, John Hancock Property and Casualty Insurance Co. sold its assets to New York-based AXA Corporate Solutions Reinsurance Group (formerly known as AXA Reinsurance Co.), which renamed it AXA Re Property and Casualty Insurance Co.

As AXA folded the new MGAs into its operation, it faced a challenge of incorporating a diverse product line into a central system: AXA Re's various MGA's market a total of 75 different insurance products, ranging from personal homeowners, marine and business owners policies.

In AXA's product distribution scheme, MGAs originate transactions. In an effort to standardize data into one repository, AXA deployed a middleware layer to enable data from multiple, disparate operating systems to be translated and then transmitted into a central data repository.

In the past, the ill-advised, if not undisciplined, approach to data consolidation might have been to stuff all the data from all the MGA affiliates into the central repository, and then backtrack to separate quality data from marginal or unusable data.

"In past experiences, I've seen data warehousing strategies that relied on 'shoehorning' as much data as possible into one administration system," says William Taylor, vice president of operations for AXA Re Property & Casualty. "That wasn't our intention. Our mission was to isolate each data element to evaluate it more effectively, with an objective to populate only quality data into the central repository. We ultimately identified minimum data specifications to create a snapshot of each policy record."

Minimum data requirements were condensed into about 20 data elements-far less than the universe of all elements held in each MGA system, Taylor reveals. "By late 2001 and early 2002, we had developed a much clearer picture of each MGA's data. As the project gathered momentum, we began to reach milestones."

AXA didn't tackle the project alone. The carrier implemented Thazar's INSsight business intelligence solution. "We wanted a structured system that could deliver customized business intelligence," Taylor explains. "This way, we can be assured of an enterprisewide view of results and opportunities. And it gave MGAs the kind of comprehensive information about their operations that they need to improve underwriting and control losses."

When insurers implement business intelligence solutions, many open up a Pandora's box of new detail, trends and revelations about both existing and potential customers.

They also discover that the power of business intelligence possesses a wide spectrum of applications and uses, which varies from, say, a policy administration system that is one-dimensional in scope and scale.

"Insurers that explore business intelligence will find over time that they'll get more than they bargained for at the time of program inception," says Chesbrough. "They're finding they can leverage business intelligence solutions to feed all sorts of business activities."

The front-end application of most automation processes is usually the catalyst for efficiency. Insurers that engage business intelligence investments realize that a Web portal is a core component of the overall solution.

The portal offers different views and correlations, reports and analytics. "Data residing in the central repository can be transmitted to the portal for users to perform data analysis. It's a point-and-click process where users can slice and dice data as they see fit," Chesbrough adds.

AXA uses an INSsight component called ProfitPortal to enable it to display and then redisplay data to affiliates, such as accounting, actuarial and underwriting units, as well as its MGAs.

"Actuaries can use data displays to look for opportunities with 'triangle' product developments," says Taylor. "The front-end tool also enables portfolio analysis and risk modeling. Each user group develops data marts to package the data they require. We also have gotten a lot of mileage out of redisplaying data to MGAs, providing customized data that we believe can help them build their business and perform optimal customer service-even look for fraud trends. We regard this aspect of INSsight to be a value-added service."

Similar to AXA Re, the Chubb Group of Insurance Cos., Warren, N.J., has identified business intelligence systems as a compelling investment to help attain its top-line and bottom-line goals. Chubb serves its customers through a network of more than 5,000 independent agents and brokers around the world.

As a reseller, Chubb is one step removed from the sale of products. However, because Chubb is not directly involved on the sales side, it has placed a greater premium on the demand to capture a complete handle on activities initiated by its MGAs.

Multi-faceted approach

In May 2001, Chubb launched a business intelligence solution called Pinpoint, which is based on a SAS-powered business intelligence and marketing application.

"Before, the details about policies, claims and the many different interactions with customers had typically been stored in many different databases. Now we are able to get better information, both from a tactical and strategic standpoint," says Jeff Hoffman, vice president of customer and market intelligence at Chubb. Hoffman led the Pinpoint effort, directing a team of experts within Chubb's Business Intelligence Office.

Chubb takes its data consolidation efforts seriously, and has recruited individuals over the years who are extremely fluent in business intelligence process maintenance. "We've made business intelligence a core competency at Chubb," says Hoffman.

While Chubb holds a great deal of internal autonomy in this process, SAS's role within Pinpoint is a crucial one. "SAS provides the back-office support that enables us to understand more about our customers," Hoffman says.

SAS's solution set is designed with various metrics that enable Chubb to use data to help foster internal efficiency and reduce losses. These days, fraud detection has become a high corporate mandate. Chubb is considering using the SAS solution set to predict potential fraud cases.

On the customer service side, Pinpoint has been designed with a metric to determine, for instance, a customer's lifetime value to Chubb, and then respond with a customized, appropriate program. Based on a specific metric developed by SAS, data is populated into respective data marts to be reviewed by specific user groups.

Retaining customers is another corporate mandate, what with the defection of a growing number of customers to competitors eager to grab higher market share. Built within Pinpoint is a customer retention index module that "reveals which customers might be inclined to let an account lapse based on similar tendencies of customers who left us in the past," Hoffman states.

Armed with this data, Chubb can confidently develop a strategy three months prior to policy renewal geared toward retaining customers who might be inclined to exit a Chubb insurance program. In prospecting for new customers and looking for opportunities to provide new services to existing ones, Chubb uses Pinpoint to feed agents quality leads on a continual basis.

"We use data mining tools to identify qualifications of existing customers and prospects. The objective is to reduce the costs of customer acquisition. We use the SAS software to establish a customer value index, differentiate between customer sets and to determine a customer's purchase propensity for specific products," Hoffman says.

In the past, an agent might flood Chubb with prospecting inquiries that ultimately turned into dead ends. Using Pinpoint, Chubb is able to give its agents a highly targeted list of prospects based on insights that have been established. "With a commercial business customer, we have tools that can analyze various cross-selling opportunities. We learn the dynamics of a customer's particular industry, capturing how many employees they have, their growth rate and their exposure to liabilities," Hoffman says.

The new approach to insurers' quest to streamline data is a stark contrast from strategies past, when insurers might have believed that no data within its many systems could be poor data. But times have changed. As Hoffman comments: "The bottom line is that, now, through our data efficiencies and changing philosophies, we strive to send out less data to our affiliates and hopefully 'hit' more."

Register or login for access to this item and much more

All Digital Insurance content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access