Three essays on Big Data Analytics, Traditional Marketing Analytics, knowledge discovery, and new product performance.

摘要

In recent years, companies have become aggressive in investing in Big Data Analytics (BDA) for marketing purposes, particularly new product development. One of the basic features of BDA is its promise in delivering automated recommendations or knowledge. For this reason, companies attempt to ascertain if BDA can improve new product performance beyond what Traditional Marketing Analytics (TMA) can.;The overarching question is whether different combinations (BDA and TMA in different levels) of analytics capabilities are able to generate different kinds of knowledge for Knowledge and Information Fusion that could improve new product performance. The aim of this study is to build and assess the Knowledge Fusion Taxonomy, and then determine the attributes that are most critical in affecting knowledge generation, Knowledge and Information Fusion, and new product development.;Multiple correspondence analysis (MCA), Fuzzy Set QCA, Partial least squares path modeling (PLS-PM), and SEM are the main statistical approaches used in this study to test the model. Heatmaps were also generated to allow users to easily explore trends or dimension patterns of items and latent variables.;In general, the study suggests that BDA is an important complementary capability instead of a competing capability with TMA. The results identified by the MCA, Fuzzy Set QCA, and PLS-PM may provide such a roadmap for firms to improve key capabilities in analytics, knowledge discovery and integration, and new product development. The study supports the hypothesized effects of competing analytics capabilities (TMA and BDA) on knowledge generation, and also a positive effect of knowledge generation and Knowledge and Information Fusion on new product performance.;In particular, both the Knowledge Fusion Taxonomy and the PLS-PM suggest that when combining information and knowledge in a complex manner, Automated Knowledge is more important than other types of knowledge. Therefore, to capture the pioneer position as shown in the Knowledge Fusion Taxonomy, companies need to build new capabilities on Automated Knowledge generation by synthesizing the unique combination of analytics capabilities. In addition, Heuristic Knowledge was also found to be a moderator when firms adopt high levels of BDA to generate Automated Knowledge. This paper establishes a solid conceptual and data analysis framework for analytics and knowledge capabilities (i.e., discovery and fusion) on new product performance. Additionally, the study provides managers a roadmap to focus on important issues in analytics and knowledge discovery for improving new product performance.