Business Information Systems development is not a easy task. Even less easy is Business Information System maintenance when the business information system isn't conceived and constructed with maintenance in mind.
Even if business information systems are easy to accomplish and maintain, developing one is a strategic activity of an enterprise. That is because business information systems are a combination of both data and process, which are at times conceived, planned, developed, and maintained with different goals, methods, and strategies. At times, data and process are at loggerheads with each other, wherein each undervalues the other.
There have been times when data modelers have declared that the data needs of a business information system developer do not exist. There have been other times when a business information system developer proclaims that data modelers are just not needed. Both assertions are wrong, and if allowed to persist will cost the enterprise way more in badly designed databases and badly engineered business information systems than would have occurred if there had been a helpful interrelationship.
Data Models, the schematic representation of stored data is the expression of an enterprise's persistently retained policy execution. Thus policy, as represented through data, can only be anecdotal if it is not formally defined through data models. In contrast, Business Information Systems are the enterprise procedures through which policy is executed and ultimately retained in a persistent state. Data Models and Business Information Systems go together as do Policies and Procedures. They are hand and glove.
Given there are significantly fewer well-engineered data models than business information systems that use data from data models, building and/or having a data model first, but certainly not in isolation, is ontologically prior to building and/or having a well-ordered set of business information systems.
The basis for this ontological priority is simple. Specifying a business information system without first having an overarching data model that acts as both a target and a constraint, produces 4.6 times more business information system work products to construct. It needs to be stated, however, that there is a synergistic relationship between the state of a data model and the unfolding and evolving state of a business information system.
Changes to the data model certainly occur during the development of the business information system. These changes are unlikely, however, to ever surpass the 4.6 work product multiplier caused by not having a well-designed data model. Thus, building the data model remains ontologically prior to building the business information system.
Because conceiving, constructing and maintaining an enterprise set of business information systems is never a small amount of money, it is important to accomplish business information development in the most efficient and effective manner practical. A key question at the outset of an business information system effort is just what should be in the business information system? Simply stated, what should be its complete functionality?
The Information Systems Plan is the key to unlocking the answer to this question. The right process content of a future Business Information System is the correctively identified set of existing Business Information Systems and Database Objects that have been assigned to a particular Resource Life Cycle Node. Once a collection of Business Information Systems and Database Objects are chosen, the overarching scope of the future Business Information System is thus constrained by the requirements of a successful Resource Life Cycle Node state transformation.
Only after the Business Information Systems and Database Objects are determined, can the database models and the business information system models finally be fully specified. At that point, the key objectives for future Business Information System development are:
- Maximize non-redundant development
- Maximize development of reusable data and processes
- Maximize ability to evolve and maintain
- Maximize Use of Metabase System
The basis for dimensioning this issue and for the computation of traditional approach costs and changed approach costs and finally the determined ROI is based on a "rule of thumb."
The size and cost of a business information system is based on the count the quantity of third normal form database tables, which is then multiplied by 80 to get the total of function points. That result is then multiplied by the function point cost, which can range from $200 to $400 (depending on the mode of traditional implementation) to get a ball-park estimate for the business information system.
Thus, if a database has 200 tables, the function points is 16,000, and the ball-park cost, at $200 per function point, the cost of a single business information system is about $3.2 million.
Many enterprises have 20 to 50 major business information systems. Thus the total implementation cost of just the business information systems inventory is about $64 million.