The cost estimate employed in the Information Systems Planning is based on a data-driven approach. That is evidenced on the fact that the quantity of business information system development function points is derived directly from the quantity of database tables: 80 function points of business information system development per database table.
Under process-driven techniques, no such metric is available because a process-driven approach has no strategy for determining a fixed work product upon which the critical business information system development steps of requirements, design, implementation, testing, and documentation begin.
Based on these two studies, and based on the multitude of studies accomplished by the United States’ Government Accountability Office, the time and resources required for process-driven approach business information system development must include work 4.6 times the number of work products than through the data driven approach.
4.1 Traditional Calculation
Given a 200 tables for a typical business information system, the quantity of Function Points at 80 function points per tables for business information system functionality is 16,000. Without a business information system generator and a cost of $200 per function point implementation, the cost for the first iteration of the business information system would be $3.2 million.
Without a business information system generator that enables completely realistic prototyping, the resulting business information system is likely to change by 25% for four cycles before quieting down. Thus the cost is another 100% for a total of $6.4 million.
If the methodology is process driven however, then the cost increases by a multiplier of 4.6. That brings the overall cost to just over $29.5 million.
4.2 Changed Approach Calculation
With the use of a business information system generator, and with the use of the data-driven approach, the costs drop dramatically. The function point cost for implementation drops to just about $50 per function point. So the initial system costs drops from $3.2 million to $800K.
There are fewer cycles of business information system evolution because the first production implementation is version 5 of 6. So, at three cycles of at most 10% changes, the additional costs are an additional 30%. That brings the cost to $1.04 million ($800K + $240K).
Finally because of the data driven approach there is no 4.6 multiplier. The final system cost is thus remains $1.04 million.
4.3 ROI Summary
For a single business information system the traditional life cycle cost is about $29.5 million. When the approach is changed to be accomplished through a business information system generator and also through the data driven approach, the cost drops to $1.04 million. That represents a ROI of 28. That is 3x a silver bullet.
Organizations typically have 100 such systems and because of the data driven approach that produces a centralized enterprise data model that becomes the common data-center of different functional subsets of these business information systems, the cost is not just a straight line savings of 100 * $1.04 million versus 100 * $29.5 million. There is an extra savings because of the ability to eliminate redundancy, conflict, and the ability to re-use large functional areas of business information system specifications because of the business information systems generator and because of the enterprise-wide Metabase System’s integrated, interoperable, and non-redundant collections of work products.
Experience with building the Metabase System shows that there is a 20% overlap across all the Metabase System modules. That would reduce the $104 million cost for 100 such systems down to about $83 million. The traditional costs would be a straight line increase because of the inability to have integration and non-redundancy. Thus the traditional cost would be about $2.95 billion.
The cost savings by the change to data driven and to business information system generation is so great as to possibly be not believable. Hence the vast majority of the 100 business information systems developed under process-driven approaches are just strung out over many man years, abandoned, or their true costs hidden.