Many data warehouse (DW) projects start with the best of intentions, that is, to get at the truth of what is going on in an insurance company. However, success has been a mixed bag with many projects only having limited usability.

Download this article that appeared last year in the Society of Actuaries CompAct newsletter.

A successful DW requires significant business knowledge covering all facets of the business.

The DW should be cross organizational and translate business knowledge into the DW rather than letting users try and apply the business knowledge in the reporting process. For example, in order to analyze policy reserves, the reserve type, reserve components and change in reserve should be readily accessible in the DW. Users will need audits to verify the validity of reserves relative to policy status and reasonableness of the reserves. They will also need to be able to analyze multiple time periods and view reserves based on either a product grouping of the business or a grouping based on the individual riders. The list of issues goes on, overlapping and integrating other business areas.

These considerations should be modeled into the design of the DW rather than relying on users to generate business rules when building reports.

A DW that does not become a true cross organizational tool serving the executive, financial, actuarial, underwriting, claims, operations and reinsurance units is not an optimized DW. An optimized DW would provide validated data and applications based on population dynamics, such as policy rollforward, reserve analysis and sources–of-earnings; specialized applications such as financial analysis, claim-lag development and expense analysis; applications that cross departments such as underwriting performance, persistency analysis and reinsurance profitability; all within the same design.