Strategic Decision-making Processes of NPD by Hybrid Classification Model Techniques

Abstract

Although a successful new product development (NPD) can yield significant profits, it is a high-expenditure and high-risk investment. Thus, the related issues of the NPD success/failure are concerned by practitioners and academicians. With this reason, the research provides a NPD decision-making method to reduce the investment risk when implementing NPD projects for companies. The research develops a hybrid classification model to use a variety of data mining classification techniques, such as Bayes Net, Lazy Learner, Bagging-Bootstrap Aggregating and Decision Trees, to help companies make decision simply and accurately. A total amount of 151 NPD data from 25 companies is collected for further empirical validation to the proposed model. The empirical results show that the critical success factors (CSFs) of influencing NPD are determined as the quantity of products, the product life cycle, the sales territory, project scale and capital. Lastly, the NPD knowledge-based decision-making system is determined based on the execution of Decision Trees. Furthermore, the internet decision-making system has been verified in this research by the use of a series of NPD data from an actual company A as an effective reference for estimating the efficiency of classifying a NPD project in the future.