The solution approach includes the following four significant changes:
- Employing a data-driven methodology versus a process-driven methodology. This is detailed during the ROI, Data-Centered Development and Management.
- The maximal use of business information system generators to eliminate coding errors. This is set out in the ROI, Efficient and Effective Information System Development.
- The deployment of the Metabase System on the critical path of work product development to capture, store, report, and evolve all SDLC work products. This is set out in the following ROIs:
- ROI, Manufacturing Integrated, Interoperable and Non Redundant Data Models, Section 2.
- ROI, Efficient and Effective Business Information System Development, Sections 2.4 and 3.4
- The modification of the first major phase to include prototyping to eliminate the vast majority of requirements and design errors and omissions.
The Whitemarsh Development Life Cycle diagram shown in Figure 3 illustrates changes resulting from this last bullet.
The changed life cycle diagram shows that the first phase consumes 33% of the first-implementation effort rather than the traditional 20%. Additionally the diagram shows that the next two phases take 66% of the effort before first the implementation is completed. This is a reduction from what was 80% to 40%. More importantly, while the traditional effort was based on just text and drawings produced from the first phase, the solution approach is based on four or so cycles of operational prototypes of the entire business information system.
The key benefit to true operational prototyping is that its four iterations almost always eliminates the vast majority of critical errors that cause system failures. The Whitemarsh Knowledge Worker Framework, shown in Table 1, maps to nearly all the work products required for business information systems. To illustrate, the first two rows of U.S. General Accountability Office’s error percent that are allocated to the Whitemarsh Knowledge Worker Framework, show that 41% of all business information system failures occur during the Scope and Business rows. This is shown in Table 2.
Another 50% of errors occur when the enterprise does not reconfigure and optimize it’s organizations and functions to take advantage of reconstituted business information systems. In short, 91% of all IT business information system failures are directly attributable to non-IT reasons.
The 41% source of errors are dramatically reduced or eliminated entirely by the Section 2 Solution Approach set out in ROI, Business Information Systems Plan.
The 50% of errors surfaced during the business information system code generation prototyping cycles provide a significant head-start on identifying the changes required by business organizations and functions to meet the needs of newly conformed business information systems. It is not that the business organizations and functions are automatically changed, however. Rather these result in the necessary lead time to accomplish these organizational and functional changes.
Prototyping business information systems is both quick and easy because the business information system’s data model will have already been developed. Developed too will be the business information system’s function model. Because of this, multiple staff can proceed in parallel to generate functional subsets of the overall business information system.
Once business information systems have been generated, they are immediately demonstrated to stake-holders with the hope and expectation that such reviews surface needed data model and business information system changes. Once these changes are quickly incorporated in the data and business information system models, the business information system can be regenerated. After four or five such cycles, the vast majority of the requirements or design omissions or mistakes will have been resolved.
The second and third phases from the Whitemarsh Life Cycle diagram start at Version 5 or 6. Once the business information system is completed, there are fewer life cycles because the first production version is Version 5 or 6. Consequently, the changes are fewer, smaller, and faster to accomplish.