How many R&D projects to develop?

The eventual success of any early research project is uncertain. It has been generally recognized that commercial and applied research projects need to be screened prior to undertaking extensive development effort. But how many projects should be continued into development? This paper presents a model to determine an average optimal number of research projects to be developed. The research and development is viewed as a two-stage process, where the task of research is to reduce the uncertainty for eventual development. Research projects consist of both technical and business evaluation, including marketing research. The number of research projects undertaken should be a function of the cost effectiveness of the research effort and the ability of the organization to support the development effort. We have created an analytical model of research and development based on the concept that the function of research is to improve the likelihood of development success. Based on this model and within its limitations, we conclude that (1) an optimum fraction of projects that should be developed usually exists and (2) that optimum fraction is critically dependent on the relative average research project cost and effectiveness compared to development.