The implementation of analytic solutions in the insurance industry has been a history of hits and misses. Emerging technologies have clashed with conservative cultures and tried-and-true methods that have worked for years. The ratio of projects started to those that are running today is estimated to be quite high, with failures far outnumbering successes.

Insurance traditionally has been a segmented industry. This exists not only in firms offering only single product lines, such as auto or life, but in large multiline companies as well, where the different business units function somewhat autonomously. While the enterprise may offer revenue targets and other goals, the sharing of data across units to facilitate cross-selling and other positive aspects of collaboration is many times given minimal attention.

Insurance used to be rather simple. People needed it. They went to an agent and bought it. Price was not an issue. Relationships and customer service were the differentiators. Business intelligence, or what passed for it at the time, also was simple. Numbers were analyzed from period to period and rates adjusted. No sweat. No stress. Computers and their resultant systems were regularly used to automate already existing processes without much thought to the overall enterprise integration and the improvement of that process. In many instances, a policyholder would be identified numerous times in life, health, auto, pension and commercial systems that could not be merged or reconciled. This redundancy of data and the replication of effort thwarted efforts to turn this data into usable management information.

Today, however, with deregulation bringing the barriers to entry down, with the Internet playing an increasingly predominant role as an information/sales channel, with numerous new product lines and financial services being offered and with the pace of society making the agent less of a major reason to go with a particular carrier, the need for more accurate and timely analysis is paramount for an insurance company's survival. But the ability for insurers to deliver this analytic capability has been less than ideal. What has been the problem? Why can't these companies with massive amounts of data turn that data into information that is sharable and useful?

In most cases, the cause can be traced to the lack of an overall business intelligence strategy. As an insurance industry CIO heading an IT department, you may have an IT strategy, a data storage strategy, a methodology and maybe even have a data warehouse strategy. You may think you have everything covered. But do you really? Have you designed a way to integrate those disparate knowledge silos? Do claims and finance know what is happening with each other's business? Are the agency force and marketing department in synch? And, as the CIO, where exactly do you fit in all of this?

Business intelligence is, in essence, a set of enabling technologies that collects and presents the disparate data into views that are intelligent and fact based. It allows for more quality decisions to be made per hour. It brings new insights and perspectives to questions and answers inherent in your data, but now moves these to a more dynamic plane. Business intelligence is not just a fad, but a real and compelling evolutionary phase of managing business as reflected in estimates that $70 billion or more will be spent by 2001 on business intelligence systems.

This new millennium will be a knowledge economy, and those insurers that enable themselves now as business intelligent enterprises will be the survivors in the future. With its massive amounts of data, ranging from basic customer information to specific information on lines of business, insurers are prime candidates for business intelligence systems that will make a difference in maintaining and growing competitive advantage. The industry needs to take a look at itself, recognize this opportunity and then adjust accordingly. Insurers must take a customer focus instead of a line-of-business focus if they are to thrive in the coming decade. The first step toward that adjustment is designing a business intelligence (BI) strategy that the firm will accept and work within.

The spirit of a BI strategy is the realization that the various business units or departments do not act in a vacuum. Data is to be shared and maximized, not hoarded and used only for a single purpose. Business rules for data elements must be consistent across the enterprise. For instance, is the definition of "policies in force" the same in every department? Once this mind- set permeates throughout the company, the implementation of the strategy is much easier. But, in the insurance industry as well as others, this is sometimes easier said than done. It takes a CIO or comparable leader with a passionate vision of this information nirvana and the strength to push through the walls that impede change to make this happen. The effort must be relentless and the message consistent.

The "guts" of the BI strategy must focus on providing solutions to the business units in a way that makes the multitude of business processes work in harmony with each other. For instance, because a BI strategy was not in place, the agency group may have developed a data mart in Oracle using one OLAP tool and the finance group may have developed another data mart using Hyperion Essbase. Both of these may have been created without the interaction of the enterprise data warehouse group (if one exists) or with the IT department at large. They may have called in consultants and built the system with their own budget. Both will likely have many data elements culled from the same sources. Ultimately, these data mart data models will not have upward mobility into the data warehouse because they were designed without the benefit of an enterprise business model. You can easily see the ramifications of separate efforts and the support, cost and data redundancy issues they pose.

Therefore, it is easy to see that a strategy for combining the common business intelligence needs of all units and departments under one group is essential. To know these needs makes it easier to provide an intelligent solution that may leverage existing BI projects or add to ones already on the board for creation. A BI strategy will detail processes for the collection of business needs and the technological alternatives for the solution. It will offer processes for the construction of the best solution and will not be mart or warehouse centric.

The strategy must contain substantive business and technological sections as seen in Figure 1.

Function Content Example
Vision An attempt to see the future and map it against the present. Why must change happen? Become more customer centric, i.e., enter new markets, establish new channels such as the Internet. Introduce new products quicker with faster ROI.
Business Intelligence Define it. Illustrate it. Describe it. An OLAP tool can drill through finance data to determine trends and spot developing situations.
E-Intelligence Define it. Synergy with business intelligence. Trends and drivers. Add profiling and clickstream data to existing BI customer data to cross-sell between lines of insurance and financial products such as annuities.
Technical Architecture Conceptual architecture to support BI. Processes and standards for determining best solution, i.e., mart or warehouse. Quality assurance, support and delivery mechanisms established. Finance needs to perform specific modeling tasks, and some of the data is in the enterprise warehouse. Do you add the missing data to the warehouse or create a mart and source it with data from the warehouse and other sources?
Organizational Who will manage the warehouse? Communicate with source systems? Will a separate business intelligence group exist with both business and IT components? How will project charters and planning get done? A business intelligence group is formed that will function as the determiner of the business intelligence best solution and establish standards and procedures.

Figure 1: Business Intelligence Strategy Areas

Other areas the BI strategy can include are security, marketing/promoting and sponsorship. Approvals by the participating business units are essential as well.

While Figure 1 is not all inclusive, it does provide a basis for the construction of a strategy. The strategy will send a message to the users that the company is endorsing the concept of business intelligence as a way to run the business. The reality of the day-to-day world, however, is that users will be more focused on attaining systems which quickly solve their short-term business problems than on strategic visioning. The challenge for the insurer is how to marry the establishment of a BI strategy that serves the enterprise while concurrently responding to the user's immediate BI needs. In other words, deliver short term, but plan long term.

There are many options the insurer might follow to respond to this challenge. Essential to going forward is the establishment of a formal business intelligence group. This organization can provide counseling, oversight and guidance on BI initiatives while concurrently crafting the strategy. It will oversee the development of immediate-need analytic solutions, such as data marts, and plan for them to ultimately serve as a foundation for an enterprise solution, such as the warehouse. In this way, the organization can phase in local solutions without committing the entire organization all at once.

For all insurers, whether single or multiline companies, whether global or small regional carriers, the establishment of a business intelligence group will enable the organization to become a coordinated and information-rich insurance company. As the industry becomes consolidated into a smaller number of major players and they establish the Web as a viable distribution medium, business intelligence takes on even more critical importance. John D. Rockefeller once said, "He who has the most information wins." By establishing business intelligence as the heart which pumps information through the insurer's body, they too, will win.

Register or login for access to this item and much more

All Information Management 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