Determinants of Business Value Realisation from the Use of Business Analytics: The role of fit, complementarity and causal ambiguity

Driven by the growing availability of data, there has been an increase in the application of business analytics to realise business value. Like any other technology, the application of business analytics is only meaningful if it leads to the derivation of business value. Prior research has highlighted the following challenges in the business realisation process: 1) inconsistencies in business value realisation, with implementations of business analytics often producing disappointing results, 2) insufficiency of business analytics to achieve business value independently, 3) uncertainty of value-creating actions based on decisions grounded in business analytics. Furthermore, the realisation of business value does not often indicate the efficacy of the process through which it is realised. Therefore, there is a need for conceptualisation, theorisation and empirical investigation to understand the determinants of business value realisation from the application of business analytics. Consequently, this dissertation’s objective is to identify, explain and characterise the determinants of business value realisation from the use of business analytics through conceptualisation and empirical investigation.

To achieve the overall aim of exploring the determinants of business value realisation, the dissertation employs an eclectic philosophical approach including the adoption of critical realism and pragmatism. The dissertation contends that business value realisation is a phenomenon best understood by using more than one theoretical perspective. Consequently, task-technology fit, complementarity and causal ambiguity are applied. Task-technology fit suggests that the best outcomes are achieved when there is a perfect match between business analytics characteristics and the characteristics of the task to which it is applied. Complementarity arises through combining business analytics and other firm resources to execute tasks or create capabilities. Causal ambiguity highlights the challenge of understanding the relationship between business analytics and business value.

This dissertation is a compilation consisting of three conceptual and two empirical publications. The conceptual publications had different goals such as explicating, envisioning and debating, which helped to characterise the determinants of business value realisation. The empirical publications are based on the qualitative research method involving data collection using semi-structured interviews. The two empirical publications had the distinct goals of theory elaboration and theory generation.

The dissertation contributes to the literature on business value by conceptualising that the better the fit (i.e. close to ideal) between business analytics and a task, the higher the business value. Fit can be improved by either modifying business analytics or adjusting tasks. However, the dissertation suggests there are limits to the modifiability of business analytics and the adjustability of tasks. Empirically, it is noted that under-fit can be persistent. When the marginal cost of enhancing business analytics starts to exceed the associated marginal business value, an organisation may stop improving the business analytics even though an ideal fit is not reached. Under-fit may diminish over time since the understanding of problems improves as they are solved. The dissertation contends that complementarities determine business value by 1) controlling the level to which the potential business value within business analytics is actualised, 2) influencing the creation or enhancement of capabilities, and 3) ensuring that task requirements are met.

Drawing on critical realism, the dissertation suggests that business analytics is an entity where causal ambiguity arises because of the resource bundle complexity of the different constituent parts. Thus, the lower the ambiguity inherent to business analytics because of its nature (i.e. characteristic ambiguity), the higher the likelihood of realising business value. Overall, the dissertation suggests that the determinants of business value are 1) the nature of fit between business analytics and tasks, 2) the capacity of the business analytics capability created from the complementarity of firm resources, and the extent to which the created business analytics capability is exercised; 3) the ambiguities of business analytics, business value and the links between them.

This dissertation contributes to the practice of how business value is achieved by highlighting the importance of combining business analytics with governance, learning and knowledge-sharing capabilities. Additionally, consideration of fit requires managers to focus on both business analytics and tasks. Specifically, not only should there be a proper application of methods and techniques within business analytics, but tasks must also be optimised. Thus, paying attention to the efficacy with which business value is derived from business analytics is important.

The dissertation suggests that future research could empirically map the changes in fit during the lifecycle of tasks to discover which task groups business analytics is an ideal fit for. Future research should also investigate the impact of interdependencies between components that constitute business analytics on the realisation of business value.