HITACHI ENABLES INSURANCE COS TO IMPROVE CUSTOMER SERVICE

The Situation

PMIC understood the critical importance of timely and accurate product performance monitoring. It had "experience study systems" designed to analyze claim costs and lapse rates. Such analysis involved comparing actual results to expected results. However, these systems had been designed many years earlier and could not provide the information currently needed. Also, they were complicated, required significant IT resources, and had to be rebuilt each time.

Furthermore, PMIC was changing its organizational structure and the way it did business. It was committed to becoming customer-centric and was implementing a customer segmentation strategy. This key initiative affected the entire enterprise. It used customer lifetime value to measure results, which demanded timely, accurate, and comprehensive experience data. Therefore, PMIC resolved to build a new Product Performance Monitoring (PPM) system.

The Challenges

PMIC was several years into the PPM project when it found the effort required more dedicated and specialized resources than planned. With other important projects, such as Customer Segmentation and Y2K, demand for IT resources was great and PMIC was unable to properly staff the project. It was particularly difficult to find IT professionals with the specialized skills required.

The HISC Solution

Recognizing that it needed help, PMIC sought to outsource the completion of its PPM system. HISC was selected based on its reputation for successful data warehousing solutions. HISC's data warehousing consultants quickly met the challenge. Through their understanding of both the business and technical issues facing PMIC, they were able to obtain commitment to the project from the project sponsor, executive vice president, and IT Group. With the involvement of key business and IT representatives, the project team solved complex project problems, accelerated progress, and completed the PPM system.

Data Warehouse Strategic Plan

The success of this effort led to HISC being engaged to develop a Data Warehouse Strategic Plan to meet the growing needs for data warehousing across PMIC. In a series of presentations to the CEO and Executive Committee, HISC explained the value of data warehousing to the insurance industry. This resulted in a highly visible strategic warehousing initiative and creation of a Data Warehousing organization within PMIC's IT Department. This organization is designed to provide ongoing support and enhancements for data warehousing. HISC also provided mentoring and skills transfer so that PMIC's Data Warehousing organization could become self-reliant.

The strategic warehousing initiative involved analysis of PMIC's business strategies and objectives. A series of facilitated Joint Application Development (JAD) sessions were held. During these sessions, data architecture and infrastructure requirements were identified, and a project plan to fulfill the long-term data warehousing needs was developed. The project plan incorporated the Hitachi "Think Global, Act Local" methodology, which calls for implementation of new functionality every 90 to 120 days within the framework of the long-term plan. Some early implementations included adding the following to the data warehouse:

• Life insurance data. This involved extraction, cleansing, and integration of data from a Vantage system and from a proprietary life system built inhouse.
• Direct-response promotion history data. A customer lifetime value project. This took advantage of the promotion history data and made the calculation of customer lifetime value faster, easier, and more precise.
• A promotion history experience study system. This provided the same level of detail for promotion experience as the PPM system provided product experience. It was needed to provide a repeatable and simplified process for the calculation of lifetime value.
• Household/customer demographic data. This provided the ability to measure marketing and product performance by external demographics.

The Benefits

With the PPM system in place, PMIC's actuaries are now able to monitor product performance on a timely and consistent basis. Changes in product performance are detected early and adjustments are made. This allows for continuous learning and significant improvement in the accuracy of product margins and customer lifetime value. In the words of PMIC's CFO, "Hitachi put us on the right track. As a result, this project was a success, and our future database projects will benefit tremendously."

The customer lifetime value project and its follow-up project to build a repeatable and more simplified process made the calculation of customer lifetime value easier, faster, and more precise. With customer demographic data available, PMIC's ability to implement its customer-centric goals and customer segmentation strategy has been greatly enhanced. PMIC can now report results by its customer segments, which are currently based on age, income, gender, and marital status. Over the long term, these segment definitions are expected to evolve as more is learned about customer preferences and behavior.

PMIC is using its enhanced customer lifetime value metrics and knowledge of customer-buying behavior to implement its customer segmentation strategy and optimize long-term profits from new business. Goals are set based on contribution to profit, not merely on production quotas. PMIC is experiencing continuous improvement in everything it does.

Technologies Utilized

The successful completion of the PPM system required implementation of new technology for PMIC, including:

• DB2 database on OS/390 mainframe
• Cobol for data extraction, transformation, and Cleansing
• Highly advanced Nucleus database residing on an NT Server to facilitate fast data query and exploration. The database utilizes full indexing, bit maps, and compression for fast execution of complex queries.
• Hitachi Data Systems 7700E storage array
• Harbor data movement software
• Brio end-user query and analysis tools

HISC is also helping PMIC's Data Warehousing organization evaluate and recommend a database for multidimensional analysis and tools for data extraction, transformation, and loading.