A methodology for assessing the effect of portfolio management on NPD performance based on Bayesian network scenarios

Firm growth and profitability come primarily from new product development. Portfolio management has been emphasized in improving new product development (NPD) performance under multiple project environments. However, few researchers have demonstrated the consequence of different combinations of portfolio management practices on NPD performance. In this study, a decision support methodology based on Bayesian network scenarios is used to simulate the effect of portfolio management on NPD performance in uncertain environments. Firstly, portfolio management factors are identified and performance criteria determined. And then, the causal relationships among the factors are modelled within similar time frames, and a Bayesian network model is developed by parameter learning from data. A case study is carried out for project/portfolio managers in Chinese firms. The most informative factors affecting NPD performance are identified by sensitive analysis, and the best and worst scenarios with different combinations of portfolio management practices are analysed. The study extends the application of Bayesian networks to assess the performance under changing conditions and highlights some managerial suggestions to improve NPD performance.