تاثیر سیاست های تاکتیکی و عملیاتی در انتخاب پرتفولیو محصولات جدید

The effective management of a new product development (NPD) pipeline is critical to guarantee the survival of the organization in the long term while maximizing the creation of value. This is a challenging goal, due to one or more of the following factors: intensive research and development investment, long and uncertain development times, low probability of technical success, and uncertain market impact and competition. In NPD management, as in any other area, decision making is commonly broken down into three independent hierarchical levels: strategic, tactical and operational; where each level uses data and models whose degree of aggregation depends on their corresponding scope and their dynamic or static character. In principle, this decomposition strategy allows managers to concentrate on the variables that are relevant to each level and therefore generate decisions that will be reflected in optimal or near optimal performances. However, there are no empirical or theoretical results reported in the literature that validate this assumption. The aim of this study is to characterize the optimality gap between the set of decisions based on a decomposition strategy and those obtained by using a comprehensive decision support approach, in which the dynamics of all the different decision making levels are considered simultaneously. For that purpose a multi-phase, multi-level Sim-Opt decision support framework capable of accommodating any set of decision making levels with any degree of detail is proposed. A specific instance of the framework is used in the context of the pharmaceutical industry to determine the effects of considering the resource allocation strategies on the composition and prioritization of an NPD portfolio. Results show that if an integrated strategy is not considered it is not even possible to roughly estimate the performance of the pipeline for the chosen composition and prioritization. The performance of the portfolios selected using a tactical decision making strategy slightly different than the one implemented in the real system proved to be significantly suboptimal, off target and sometimes unreachable.

New product portfolio management is a key part of the dynamic decision process used by senior management to operationalize a business strategy (Roussel, Saad, & Erickson, 1991). It specifically delimits the future directionality of the enterprise by selecting and prioritizing the development of a limited set of products using the resources available. Therefore, the efficient and effective management of such a portfolio is critical to guarantee the ability of the organization to compete and survive in any possible future scenario. However, this is a challenging problem due to the characteristics of the development pipeline, namely, (1) the presence of uncertainty (internal and external), (2) the interdependency between projects, (3) the limited availability of resources, (4) the discrepancy between strategic and tactical/operational goals, (5) the dynamic propagation of the effects of lower level (tactical and operational) decisions in response to the resolution of uncertainties, (6) the overwhelming number of decision and state variables due to the length of the time horizon and the combinatorial nature of a portfolio (Blau, Pekny, Varma, & Bunch, 2004), and (7) the difficulty to terminate projects once they are begun (Buyukozkan & Feyzioglu, 2004). A wide variety of strategies capable of dealing with one or more of the features mentioned above have been proposed to help manage new product development portfolios (Cooper, Edgett, & Kleinschmidt, 1999; Hans, Herroelen, Leus, & Wullink, 2007). However, all of them assume that the decision making process can be decomposed into independent hierarchical levels, and therefore use models that do not take into consideration the dynamic propagation of the effects of lower level (tactical and operational) decisions when projects are analyzed as part of a portfolio. This assumption led to the development of decision support methodologies based on aggregated models with some type of static or semi-dynamic linkage that are valid when the process is close to deterministic, there is significant excess capacity at every stage of the process, or the pipeline does not contain multiple projects. However, in the highly uncertain, constrained, and dynamic environment of R&D the effectiveness of such methodologies is not so clear (Anderson & Joglekar, 2005). Decisions at any level in the pipeline may propagate in unpredictable ways, seriously affecting the performance of the portfolio in the long term. Such concerns are usually addressed through the use of sensitivity analysis and the examination of multiple scenarios. The problem with these techniques is that resource constraints, project interactions, and dynamics created by lower level decision making strategies may force the system to behave in ways that can not be captured in the static open loop models commonly used at the strategic level, creating a false perception of the capability of the pipeline. Thus, it is critical to determine if it is necessary to use a comprehensive approach to optimize the portfolio, in which the dynamics of all the different decision making levels are considered simultaneously. This work explores the implications of the decision making strategy at the tactical level on the selection and prioritization of new products using the pharmaceutical industry as case study. For this purpose a quantitative decision support framework capable of accommodating all of the characteristics of the pipeline with any degree of detail is proposed and implemented. The rest of the paper is organized as follows. Section 2 reviews the literature that is relevant to the problem. Section 3 outlines the problem through simple examples. The proposed decision support framework is presented in Section 4. Section 5 presents a specific realization of the framework and describes the case study used to explore the implications of using different strategies to allocate resources in the selection and prioritization of projects in an R&D development pipeline in the context of a pharmaceutical company. Section 6 provides results and discussion for the case study. Finally, concluding remarks and perspectives are presented in Section 7.

All the decision support methodologies found in the literature are based on decomposition and model simplification assumptions, which are typically not validated. A multi-phase, multi-level Sim-Opt framework that does not rely on such an assumption was proposed, and a specific realization of the framework was implemented in the context of pharmaceutical R&D portfolio. The study demonstrated the limitations of the decomposition and model simplification based approaches. It was found that by decomposing the decision making process the strategic decisions obtained generate significantly suboptimal, off target and sometimes unreachable performances. It is relevant to note that the realization of the framework used to reveal these pitfalls is at the same time a deployable methodology to help decision makers to better manage their development pipeline and better understand how interactions, uncertainties and decision making strategies affect the pipeline performance. However, there remains further work to be done to achieve improvements in three different areas: reward-risk characterization, simulation fidelity and speed. With respect to the first limitation, it is clear that although the use of the probability of losing money is a common measure of risk in portfolio management, it is an incomplete measure. It does not capture the magnitude of the losses and the variability around the EPNPV. We recognized that situation and decided to explore different combinations of the three types of risk. In the process we realized that that the challenge of selecting a risk measure and a return measure as well is substantially more involving from a theoretical perspective. The problem is due to the fact that the state of the system, and the flexibility to allocate different levels of resources and terminate projects determine at every point in time the level of risk that is being faced, thus, requiring a matching discounting factor to calculate the ENPV. We anticipate that the use of real options is the most promising approach to solve this problem. However, the Sim-Opt approach and real options do not mix easily. Future research will be focused on bridging this gap. The second limitation, simulation fidelity, can be seen in two places: The three mode optimizer in the inner loop does not really reflect the infinite set of options managers have in real life, and second the market interactions were represented as simple statistical distributions. In order to address the first part of this issue we are currently working on the incorporation of “continuous” resource allocation models, while for the second part the use of agent based simulation is being explored. The third limitation is the standard pitfall of all simulation-based methodologies, namely, computational intensiveness. Ideas based on ranking and selection are being considered to mitigate this issue to some extend. To conclude, the proposed framework is just a piece of the enterprise wide optimization puzzle. As a first step, future research is required to integrate it to the corporate strategy for the “external” development pipeline (e.g. in licensing, joint ventures, acquisitions, etc.), and to model the correlation between the pipeline performance and the creation of value for stockholders.