Selecting An Application

 

It  is desirable to have all new systems suggested and investigated at one point of time so all possibilities can be considered and a subset selected for implementation. Unfortunately, ideas for systems arise almost at random. Some length of time is required to study the suggestions before a decision is made on whether to undertake a suggested application. The decision process recommended below concentrates on the selection of an alternative for a single proposed application. It does not attempt to evaluate an entire portfolio of projects because we are rarely in a position to compare the entire set of proposed projects at one point of time.

 

Providing Decision Information: The approach suggested here can be used for decisions at either the preliminary survey stage or the feasibility study stage. The major difference between the two is that the feasibility study contains more data than the preliminary survey and presents more-refined cost estimates.

 

The first task of the selection committee is to agree with the information services department on the number of alternatives for a single project and how the alternatives should be developed. As an example, suppose one user department has proposed an inventory control system. The alternative might include (1) doing nothing, (2) purchasing a packaged program from a computer services vendor, and (3) building an on-line system. Each of these alternatives for an inventory control system meets some percentage of user needs at different costs. Probably three or five alternatives for each proposed application are sufficient. However, there should always be more than one alternative for a new system. Selecting the first alternative (doing nothing) is equivalent to deciding that a new system is infeasible.

 

The next step is for the committee to agree on a set of criteria to be used by the information services department in evaluating each alternative. The set of criteria should be as complete as possible so that no important evaluation factors will be overlooked. However, the selection committee should avoid enumerating too many criteria, or the data collection and processing requirements for evaluation will become a burden. Each criterion should be measured on a common scale of, say, 1 to 7.

 

Making the Decision: If the recommendations above are followed, the selection committee should be in a position to review a series of alternatives for each application proposed for implementation. Each alternative should be evaluated on the criteria established by the committee. In this hypothetical problems, the selection committee is considering either an applications package or a custom-developed inventory system.