Parametric cost models are powerful tools that can be beneficial throughout the three major areas of a development program: planning, estimating, and control. Parametric cost models can be used to generate program cost estimates at a conceptual level and support a bid/no-bid decision in the planning stages of a proposal. They can be used to estimate program costs at a detailed level for submission in a proposal or help substantiate estimates derived by other methods. They can also be used as a measurement tool to improve quality, control costs, and identify risks during contract performance.

Processes that exist for these cost-related activities at IBM Federal Systems Company (FSC) in Manassas, Virginia are discussed in terms of their limitations. The use of parametric cost analysis can supplement the current processes and provide structure and objectivity where little are evident. It can also reduce time, cost, errors, and labor involved in performing these activities. The result of the paper is a re-engineered proposal process incorporating the use of parametric cost models.

This paper presents a conceptual discussion. Limited implementation is described in terms of an example program, Beta, where a parametric cost estimating method is tested. Based on the results of this example, a quantitative evaluation is made. Further plans for validating the proposed solution and justifying its use are outlined. Common concerns and objections about parametric cost models are addressed. The Parametric Review of Information for Costing and Evaluation of Hardware (PRICE-H) model is used as a case study to illustrate some of the possible applications of parametric models and how they are performed. Specific limitations of the PRICE model are discussed.