مدل انتخاب گزینه های R & D برای تصمیمات سرمایه گذاری

Technology centered organisations must be able to identify promising new products or process improvements at an early stage so that the necessary resources can be allocated to those activities. It is essential to invest in targeted R&D projects as opposed to a wide range of ideas so that resources can be focused on successful outcomes. Typically, a number of options and tradeoffs are encountered; the selection of the most appropriate projects is the aim of R&D selection models. Although capital budgeting and financial portfolio management offer a similar style approach, the techniques used for the solution of those is different to that used for R&D project selection. The reasons for this are that project selection is complicated by many factors, such as uncertainty, interrelationships between projects, changes over time and success factors that are difficult to measure. Thus, a mathematical optimisation approach in isolation is not practical. Project selection models not only have to consider these problems but also that there are different types of R&D. The spectrum of R&D ranges from low budget exploratory research to large budget product development. This paper reviews the development of a project selection and evaluation tool that can be applied to a wide range of research, technology and investment decisions. Firstly, the background on project selection models is given. This is followed by the introduction of the model and its application to a sample group of projects. Finally, some conclusions are discussed as to the applicability of such models.

Most organisations use some form of project selection/justification tool, which, according to Liberatore (1987) is usually one of the standard financial analysis methods, such as cost–benefit analysis or discounted cash flow. Techniques such as mathematical programming are not widely used due to the diverse nature of the projects. A wide variety of project selection models have been assembled over the years, including linear programming, scoring models and checklists. Many early selection models were based on an optimisation approach. Given a number of projects and a pool of resources, the portfolio of projects was optimised to a certain criterion. This usually involved the conversion of the attributes of a project into a single monetary value. There is little information on the application of these early models to project selection decisions. For R&D type selection decisions, Moore and Baker (1969) suggested that the models were not entirely suitable due to a lack of input data. The complexity of the models and the problems of application can be a deterrent. The fact that models were not being used was summed up by Moore and Baker (1969): “Management is not likely to use any model in deciding between projects, the use lies in the range of information generated for making selection decisions”. In other words, the process of gathering the information was perceived as the main function of the model. Schmidt and Freeland (1992) describe traditional optimisation processes as “classical” models, where the focus is on the outcome. “Decision event” or “systems” models focus on the process by which the outcome is reached. The result of a systems approach is that the information generated in applying the model is used in the subsequent decision making. An improvement on the classical type of model, and a direction of subsequent research, was “Multiattribute Utility Theory” (MAUT), of which an overview is given by Pearson et al. (1996). Individual projects are rated on different merits using processes such as scoring models or checklists. Rather than having a single constraint to which the project portfolio is optimised, MAUT encompasses a number of different areas, such as risk, reward and resource allocation amongst others. Recently, studies of applications of models have concluded that the models do work. A survey by Cooper et al. (2001) on the uses of portfolio management models concluded that using some form of portfolio selection tool or system is greatly beneficial. These benefits included a better balanced and aligned portfolio. Farrukh et al. (2000) present a workshop approach to the problem of the design and application of a selection method. The involvement of multidisciplinary teams in the sessions enables a broad view of the problem and the application of more than one evaluation method. Findings by Cooper et al. (2001) showed that those that use more than one selection method have the best results, since no single method has the best attributes in all areas. It appears that the trend in applying selection models is to move away from the application of a single method and to move towards a composite approach of using a number of selection methods. Those criteria that produce effective programmes involve information on markets, customer needs, competitors and regulatory and environmental concerns (Adler et al., 1992). Incorporation of these factors into a selection method hints at the use of a composite model. Pearson et al. (1996) argue that the process of project selection has given way to project evaluation. Rather than having an R&D budget and a pool of projects, the situation has become one of ensuring that the corporate requirements of the R&D function are met. This not only includes the initial evaluation of projects, but also continual review throughout their lifespan. The Stage Gate™ system of Cooper (1986) is a way to provide a frame for the constant evaluation of projects from initial idea to product launch. The idea is that at each stage of a project’s life it is assessed. This can be in the form of a go/kill decision. This stage gate process was expanded by Cohen et al. (1998) to include basic research projects by adding an extra gate to the beginning of the process. A scoring model was used as the method of assessment.

The model has been applied to a group of R&D projects to provide a basis upon which to evaluate it. This has consisted of the processes of designing the stages, gathering the required information and applying the stages. The intention of the model is to compare like with like and be applied to different types of R&D projects. Generally, within large organisations there are formal decision making processes, which, for the selection of major investment projects, usually involve undertaking a business case study. Using the business case as a lever, the management can decide on the best course. The range of information contained in the business case is vastly more than can be incorporated into a selection model. Despite this, for small to medium projects, there is some value in using a selection model. This value is summed up below. What it does do: • Provides insight into the selection process—an indication of those metrics that influence the decisions is gained. • Enables a “systems” approach—the application of the various stages has produced a wide range of information at many levels. • Gives justification to the choice of a particular project and records the results of decisions made along the way. • Increases the amount of information available to the decision maker. What it does not do: • Give an optimum project solution—the range of input projects is very diverse, making an optimisation process impractical. • Completely remove the manager’s decision—an element of human input is needed to keep watch over the procedure. 6.1. Incorporation into the stage gate process The stage gate process of project selection was introduced earlier. One aspect of it is that as projects progress through the review process, they change in nature from being basic research projects to experimental development projects. Initial high risk, low reward projects become low risk higher reward projects by the later stages. Rather than being an alternative to the stage gate process, the model presented here lends itself to incorporation into each gate review and provides the method of assessment for making go/kill decisions. An alternative method, presented by Cooper et al. (1997), uses a portfolio review alongside the stage gate process. The portfolio review is used to assess the entire portfolio of the company to ensure that the correct mix of projects is present. 6.2. Conclusions The benefits of using the model presented here are a formal decision structure and communication of information about projects. The additional benefit of extending the process to include basic research is gained. The application of such a model does require more work, and hence more cost, as although the information needed may exist, it must be collected and arranged. From this study, it appears that the role of the selection model is more relevant to lower budget R&D projects, where little or no structured decision making is used. This might be the evaluation of basic research projects within large companies, for example. Through the concurrent application of several types of selection methods, business strategy in addition to financial methods aids the decision making process. The model allows a step-by-step approach incorporating a number of sources of information. Although all of the individual pieces of the model are in use and exist as selection methods in their own right, bringing them together in a framework allows the positive features of each to be used. It is suggested that models such as this be applied using this framework approach which allows flexibility for selection decisions within different industries.