Reducing duration in new product development through project design models with real world constraints including portfolio planning

Abstract

It is typical of medium to large sized organizations involved in new product development to have a portfolio of projects at all times. The planning for the execution of these projects is usually done at an individual project level using traditional project planning tools that were developed in the early half of the 20th century to handle work in factories, with repeated standard work tasks that take up a fixed duration and sequence. However, the work performed in these knowledge intensive organizations is ill suited to be planned in a sequential fashion with fixed resources without taking into account iterations of work. The tasks in new product development projects tend to be iterative owing to concurrent dependencies between them, the teams on these projects tend to be geographically far flung leading to coordination and other socio-technical challenges. At the same time, the complexity of projects tends to grow as organizations deliver more system level solutions. These issues of iteration, concurrency, team distribution and complexity lead to an increase in demand for coordination, which is a real effort with real costs. Organizations face the challenge of making good forecasts that can include these real world execution issues through their project plans. Can project plans that include these in their planning forecast outcomes better? As the projects grow larger in size and count, and when the organizational budgets and practices do not keep pace with these changes, organizations tend to face resource constraints. Can the challenges of iteration, concurrency, distribution, and complexity be addressed by viewing the planning process at a higher level? That is, by conducting the process of planning at the level of a portfolio of projects versus planning for several individual projects? This work addresses these real world challenges in the planning of projects and demonstrates the efficacy of including concurrency, distribution, complexity and iteration in the project planning models using data from two large projects in a semiconductor organization. Analysis reveals that these models produce more insightful forecasts than the models using traditional tools of forecasting. This work addresses some of the impediments that come in the way when addressing these concerns. A project portfolio model based on these two projects demonstrates that such a model helps in better resource sharing and forecasts a 32% improvement in the duration of one of the projects compared to the forecast based on planning for the projects individually. The work also demonstrates the concept of design of a project portfolio and illustrates how a portfolio of projects model allows new product development organizations to create flexible project plans that handle real world constraints leading to a reduction in the project durations of the projects in the portfolio.