Technology portfolio management: optimizing interdependent projects over multiple time periods

Many product development methodologies strategically allocate resources to create a portfolio of technology projects. These methodologies work well for projects that are predominantly independent of one another and occur in a common funding cycle. When the projects become interdependent, the complexity of optimizing even a moderate number of projects over a small number of objectives and constraints can become overwhelming. This paper presents an approach to account for and quantify the interdependencies of technology projects and explains how it was integrated into a non-linear, integer program model at The Boeing Company. The model evaluates if and when to start funding a project over a four year period. The effectiveness of each project in each year is influenced by whether or not the projects that it is dependent upon are also funded in that year. The funding choices influence the balance of identified risk in the portfolio, the support for designated categories of objectives, the cost and the estimated financial return of the portfolio. Optimizing the model identifies the funding strategy that maximizes the potential return subject to the spending and portfolio balance constraints. Once the "optimum" strategy is found, the model evaluates the impact of alternate portfolio strategies defined by the user. Thesis Advisors: Professor Stephen C. Graves, Sloan School of Management Assistant Professor Anna C. Thornton, Department of Mechanical Engineering