Summary: GAO was asked to study how conversion costs affect the procurement of new computer systems. Of particular concern was whether including conversion costs in computer procurements would result in selecting the system that would cost the Government the least over the life of the system. Conversion costs, a significant factor in computer replacement procurements, may run into millions of dollars when noncompatible equipment is acquired. Therefore, an agency can avoid these costs by staying within the incumbent vendor's product line or acquiring compatible equipment. In this study, six computer procurements were reviewed in which a different make of computer was acquired. The primary objectives were to ascertain whether conversion costs were considered when selecting the successful offerors and whether the seleted proposal cost the least over the life of the system.
Conversion costs varied significantly in the procurements studied. Each installation had underestimated the cost and time it would take to convert its application software to the replacement computer system. Currently, no official written procurement policy or regulation deals adequately with the treatment of conversion costs in evaluating proposals. While regulations establish policies for the management acquisition, and utilization of automatic data processing (ADP) equipment and recognize conversion as a cost factor, they only provide unclear general guidance subject to misinterpretation. According to a draft guideline on how to handle agency procurement requests involving augmenting or replacing an agency's ADP equipment, most conversion costs should be included when evaluating vendor proposals to determine the lowest overall cost. Two other major problems to be considered in conversion are: estimating costs with reasonable accuracy and managing the conversion process. The development of techniques for better estimation of conversion costs is among the objectives of the Federal Conversion Support Center. Actions to be taken to ease conversion include: advance planning; developing, adopting, and enforcing standard documentation and programming techniques; identifying problem application areas for redesign; setting up and maintaining comprehensive source program and test data libraries; and developing and maintaining an inventory describing the characteristics of existing application software.
