Ridderprint

Chapter 2 40 BDA. The co-citation network (Figure 2.1) – exploring the intellectual roots of the BDA- performance debate – demonstrated a strong divide between the core BDA research stream and the clusters developing and implementing predictive algorithms. Similarly, the historiography (Figure 2.2) – exploring the historical evolution – and the bibliograhic coupling (Figure 2.3) – exploring the future evolution and trends – illustrated the weak overlap in the shared knowledge and discourse between the research streams covering strategical issues in BDA research (e.g., value, management, ethics) and those covering operational implementations (e.g., algorithms, applied analytics, predictive analytics, text analytics, clustering). Relatedly, it is worth to note that over a third of the primary documents could not even be included in the bibliographic coupling analysis because they lacked bibliographic connections to any other document in the network. This is a worrying development, considering that a vast amount of information and knowledge – including potential best practices or novel algorithms – is not diffused in the greater scientific community. Fortunately, this seems to be improving. Our historiography (Figure 2.2) demonstrated that the first bridges between these two research streams have recently been established by Ballings and Poel (2012) and Manyika et al. (2011). Third, our studies suggest that the various management functions in organizations are in different stages of BDA maturity. In particular, the use of BDA seems established in relation to financial management and customer management and development, where big data and the more advanced statistical algorithms are already widely researched, discussed, and applied. Figures 2.1 and 2.3 suggested that developments within marketing, supply chain, and IT are on their way as well. However, research in these functional domains is focused mostly on the high-level, strategic impact of BDA (Chen et al., 2015; Germann, Lilien, & Rangaswamy, 2013; Trainor, Andzulis, Rapp, & Agnihotri, 2014; Trkman et al., 2010) rather than actual applications or individual-level predictions within these areas (for some exceptions see Ballings, Van den Poel, Hespeels, & Gryp, 2015; Chi, Ersoy, Moskowitz, & Ward, 2007; Esfahanipour & Mousavi, 2011). The other management functions seem to be trailing behind. For instance, although studies mention the rise of BDA and algorithmic intelligence in the HR domain (e.g., LaValle et al., 2011), little focused academic research has been conducted in this space. Arguably, this is undesirable: HR missing the big data bandwagon may imply a loss for organizations and cause harm for employees, whose interests would consequently be overlooked in BDA initiatives (Angrave et al., 2016). Similarly, we did not encounter studies on the use of BDA in the public sector or in relation to legal, procurement, M&A, health and safety, or facility management, leaving potential impact for predictive analytics and data-driven strategies in these areas (cf. Reinmoeller & Ansari, 2016; Sheng et al., 2017). A third insight relates to the cluster on the corporate social responsibility that arose in both the co-citation and bibliographic coupling networks. Although the core publications in these clusters did consider the effect of (perceived) corporate social responsibility on organizational performance, they may have had little to do with BDA (e.g., Chatterji et al., 2009; Lucas & Noordewier, 2016; Waddock & Graves, 1997). On the one hand, due to the proprietary nature of social and environmental ratings such as those of Kinder, Lyndenberg, Domini Research and Analytics (currently MSCI), we cannot