In the pantheon of famous last words, which includes Julius Caesar asking, "Et tu, Brute?" and Civil War General John Sedgwick claiming, "They couldn't hit an elephant at this distance" just before being killed by a Confederate sharpshooter, we need to add a new entry: "A data warehouse will solve all of our problems,"--former IT director.

Maintaining an effective connection between detailed data (entries, documents, transactions) and end reporting is the ongoing quest of any management accountant. Enterprise resource planning (ERP) systems or implementing data warehouses often have been touted as the "silver bullet" for an organization's information problems, but these not-so-easy "easy" solutions fall short of solving an organization's real problems.

The promise of ERP systems to provide a centralized, all-inclusive solution for all data and processes within an organization aims to solve the disconnect at the root. The failure to live up to this promise is evident from a number of indicators present in most, if not all, information systems.

* Different information systems are still used in subsidiaries and peripheral units.

* Older applications are used for specific needs that aren't covered as effectively by the main ERP system.

* The ever-present proliferation of official and unofficial spreadsheets used to provide different views of data for specific purposes and reconciliation continues unabated.

Indirect proof of the failure of the ERP promise also lies in the widespread use of ETL (Extract, Transform, Load) and related business intelligence applications. These are based on the concept of a central data repository to which all corporate data, whether residing in the main system or in peripheral, purpose-specific applications, is transferred for consolidation, analysis, and end-reporting purposes.

Do these solutions really add value to the business reporting supply chain, or do they create more issues than they actually solve? By imposing an intermediary data warehouse between the source data and the end reporting, these applications also impose predetermined rules and dimensions to drive the data "transformation" (the "T" in "ETL"). Does this cause an even greater disconnect between source data and end reporting? Are there alternative approaches that can meet the same goal with additional overall advantages?