Putting It All Together.

Despite false starts and lackluster results in many early initiatives, data warehousing is finally becoming a mainstream technology in the insurance industry.

The challenges of mounting competition, shrinking margins and flat premium growth have created a call for action in the insurance industry, resulting in a refocus on the use of technology to remain profitable and competitive. Perhaps more than any other challenge, the emergence of convenience channels, such as the Internet, has generated a new sense of urgency among insurers to protect their existing client base while trying to capture a larger share of a limited premium pool.

The answers to this challenge are locked away in the currently inaccessible wealth of customer data, the byproduct of years and years of insurance transactions. The key to unlocking that data and putting it to work for the business is found in technology in the form of data warehousing and business intelligence.

Early on, the expectations for data warehousing technology in insurance were higher than in most other industries because of the informational nature of the insurance industry. After all, the value and success of the business of accepting risk is based primarily on complete, accurate and timely information. Yet, of the approximately 3,300 property/casualty insurance companies in the United States, only a handful have reaped the benefits of successful data warehouse implementations. For each early data warehousing success, there has been at least one failure. By some estimates, 60% of early data warehousing initiatives failed, according to the Gartner Group.

These early failures were not due to the technology, but rather to a lethal mix of poorly set expectations, inadequate planning assumptions and lack of in-house expertise to implement data warehousing effectively. While this may sound discouraging, these are the typical lessons learned when a new technology is adopted.

On the other hand, those in the insurance industry that have had the resources, time and expertise to invest have seen impressive results: in one case cutting underwriting time in half, in another saving $3 million to $4 million in data acquisition costs. (See "Cutting Through the Confusion," page 122).

No Assembly Required

Despite the false starts and lackluster results, the industry is finally poised and ready to accept data warehousing and business intelligence as part of its mainstream information-technology activities. However, this acceptance is not unconditional and requires the data warehousing and business intelligence vendors to offer complete, ready-to-deploy insurance solutions rather than the early build-your-own tool sets. The insurance industry is no longer willing to accept the early data warehousing offerings with long implementation cycles and multimillion-dollar investments required to achieve any business benefits. As access to information becomes increasingly critical for insurers, the need to implement a solution becomes urgent, dictating a need for rapid implementation and quick delivery of business value.

Budgets for centralized data warehousing software in the insurance industry have decreased as much as 60%, according to a Meta Group study. At the same time, however, spending for data marts and data quality has increased by 50%, according to the same study, Data Warehouse Marketing Trends/Opportunities. These findings signal that the insurance industry has not slowed down its adoption of data warehousing and business intelligence, but it has shifted the focus from the all-encompassing enterprise model to a more realistic and manageable approach of data cleansing and conversion combined with specialized data warehouse solutions and prepackaged systems.

Upping the Ante

The insurance industry spends an estimated 3.3% to 3.4% of premiums on technology compared with banking and financial-services firms which invest an average of 9.2% of revenues in information technology according to Conning & Co. and Deloitte & Touche.

Although size projections for the overall data warehousing market vary, growth estimates remain strong, ranging from 35% to 40% annual compound growth rate. Meta Group estimates that the data warehousing market will reach approximately $15 billion this year. While the insurance industry represents only a portion of this investment volume, using a conservative 6% market share translates into a $900 million investment in data warehousing and business intelligence in insurance in the coming year, according to Gartner Group and Dataquest's report, Vertical Market Opportunities in Insurance.

The real key to data warehousing investment in the insurance industry is management's ability to change its view of technology from an expense and necessary evil to a strategic tool. "To effectively compete with these banks, mutual funds, and securities firms, insurance companies will have to invest more in technology and do a better job analyzing customer preferences if they are not to lose share to others," according to the Deloitte & Touche report, The Global Insurance Market: Surviving in a Competitive Environment.

The pace of change will accelerate as more insurers make the transition "from operational, expense-driven use of technology to proactive, strategic use," according to the Conning & Co. report, Insurance and Technology, Leveraging the Drivers of Competitive Advantage. "Those that can effectively make the transition will maximize their success."

Forging Ahead

False starts and failures marred the insurance industry's early experience with data warehousing and business intelligence. Lack of in-house data warehousing expertise and skills contributed to the situation. Y2K issues provided a real, yet convenient, delay But successful early adopters derived benefits from their implementations of data warehousing and business intelligence solutions.

The lure of these potential benefits, in an era of flat premium growth, rising operational costs and intensified competitive environment, explains the insurance industry's growing interest and increased investment in data warehousing and business intelligence technologies. Nevertheless, their successful implementation remains highly dependent upon two critical elements:

* shifting from technology-as-an-expense mentality to technology as an investment; and

* the ability of data warehousing vendors to deliver insurance-specific and easy-to-deploy complete solutions rather than the generic data warehousing tool sets that were the only "solutions" available until recently.

Crispin Burdett based in Dublin, Ireland, is commercial director of Olympic Amis.

                       Cutting Through the Confusion
                 Examples of data warehousing results for
                           different companies.
Underwriting Time          Underwriting time cut in half, with
                           quotes for auto
                           insurance delivered within 24 to 48
                           hours, and policies
                           issued within three to five business
                           days.
Data Quality               Company now guarantees that the rate
                           the agent
                           quotes is the rate at which the policy
                           will be written.
Cost of Data Acquisition   Reduced by $3 million to $4 million
                           per year by eliminating
                           redundant data. For example, state
                           motor vehicle
                           registration data had previously been
                           purchased from
                           as many as six to eight sources in
                           each regional office.
Improved Retention Rates   Augmented internal data with lifestyle
                           data from
                           external sources. Retention rates
                           improved by 0.5%.
Fraud Detection            Generated multimillion dollars of
                           fraudulent-claim
                           recoveries and improved claims
                           investigation personnel
                           productivity by 20% to 30%.
Increased Cross-Selling    Database marketing to identify
                           desirable customers to
                           Whom products should be cross-sold
                           resulted in
                           customers owning an average of 4.5 of
                           its products.
Enchanced Direct Marketing 287% improvement in marketing
                           effectiveness after
                           database marketing program
                           restructured.
Source: Conning & Co.