PROJECT SELECTION

Selection of projects for implementation must be performed in consideration of the organization’s needs and wants, corporate strategic plans, realistic expectations for sophistication of the deliverables, success attributes of the project, and the constraints for the project’s success. To make consistent and logical decisions in prioritizing and selecting projects, a company-specific process of evaluating projects must be established.

Ranking of projects is most commonly conducted with the use of an index, sometimes called a metric, or a group of indices called a model. In general, models are easy-to-use characterizations of operations, organizations, and relationships. The models range from very simple to very complex, although all models, even the very sophisticated ones, are only a partial representation of the reality that they attempt to portray.

Although models are very useful as aids in the decision-making process, they do not fully duplicate all facets of the real world. On the other hand, with this focused feature of models in mind, models can be exploited to purge extraneous elements of a problem and to highlight the important elements of the issues surrounding project implementation. The objective of the project selection process is to rank the projects for implementation in the organization’s best interest. The selection process, therefore, would use a model that is specifically formulated to optimize parameters such as organizational goals, project cost, project duration, and the anticipated attributes of the project deliverable.

The numeric results of the model must always be tempered by the experience and professional expertise of the organizational entity charged with managing the project portfolio. Ideally, this entity is the project management office, although the function might be performed by an ad-hoc, or standing, team designated by upper management. The project portfolio management group is expected to make a judgment on project viability using information generated by the model, and in consideration of the model’s limitations and constraints and/or its constituent indices. This judgment, however, will be partly predicated on the quantified rankings provided by the models or indices.

During a typical project selection process, it is necessary to identify organizational priorities for projects, establish metrics to evaluate project objectives in the light of these priorities, and formulate company-specific scoring schemas for the project ranking process. There are four sets of basic elements for the selection process: organizational priorities, project objectives, the selection model, and the ranking process (see Figure 1-1).

Indices used for project selection tend to fall into two major categories. The first category includes quantitative indices that are generally based on financial characteristics such as:

Total cost

Cash flow demand

FIGURE 1-1 Project Selection

Cost-benefit ratio

Payback period

Average internal rate of return

Net present value.

The second category includes qualitative indices that are intended to measure subjective issues, such as operational necessity, competitive necessity, product line extension, market constraints, desirability, recognition, and success (see Figure 1-2).

A project selection model is very organization-specific and should use a customized combination of indices to satisfy the organizational project selection objectives. Since models depend on numeric input from the indices, even subjective indices ultimately will have to be quantified. Accordingly, a measure of refinement will be added to the selection process if the quantification of the subjective model indices is based on consistent and organization-specific procedures. Then, using the project data, organizational priorities, customized model, and a consistent scoring process, it is possible to systematically, and in a formalized fashion, rank the prospective projects.

FIGURE 1-2 Qualitative Indices

Figure 1-3 shows a graphical depiction of a selection model that is composed of four indices. The simple summation or weighted summation of the results of the indices is used to compute the rating of the project by this model. Figure 1-4 shows a sample of a customized model. The possible total points for a project are 100 and, therefore, prospective projects can be compared on the basis of the total score yielded by this model. The importance of each of the indices in this model is signified by the points assigned to each index, which has been customized for a particular organization. The relative importance of these indices is reflected by how many maximum possible points are assigned to each index.

To extend the illustration a bit further, a hypothetical project might receive a total score of 60 points on the basis of the scoring that the project received on individual indices that were identified and ranked specifically for that organization (see Figure 1-5). It should be noted that the score of an individual project is not as important as the relative scores of several projects that were ranked using the same model and the same rating process.

When conducting project comparisons, it is useful and logical to normalize the base of reference. One of the more convenient techniques is to use the direct cost as the base of reference for the prospective projects. Compartmentalizing the direct cost of the project from other elements has the advantage of developing a project estimate strictly from the standpoint of the level of effort. Other elements such as indirect costs, overhead, and return on investment are important values; however, since they tend to be somewhat variable with organizational entity and time, they might add unnecessary inaccuracy to the project selection process.

FIGURE 1-3 Project Selection Model

FIGURE 1-4 100 Point Project Scoring System—Maximum Points Possible

FIGURE 1-5 100 Point Project Scording System—Total Project Score