A Tool To Assess The Success Of Business Intelligence Implementation Within Free State Government Departments: Task Technology Fit Perspective

Abstract

This study employed an information systems (IS) framework development based on a case study. Data collected using a questionnaire was analysed using factor analysis technique to examine the effectiveness of the widely used Vulindela System (VS) – a business intelligence (BI) technology – in the public sector, with special focus on three service-delivery-oriented government departments, namely the Treasury, Health and CoGTA in the Free State Province, South Africa. Two principal component analysis (PCA) models were built to extract latent factors capturing the weakness and strength of the VS, with the first model assessing the internal consistency and adequacy of the survey questions to measure the VS as an observable construct, and the second model is estimated based on the 12-Task Technology Fit (TTF) evaluation theoretical framework proposed by Goodhue (1995). Firstly, the evidence from the survey reveals that sizeable number of main users (about 70%) of the VS perceived the system as a: user-friendly-web-based IT system, an easily accessible, problem-solving and flexible BI tool capable of executing analytical (financial) and decision-making related tasks, amongst others. The effectiveness of the System is, however, constrained by lack of technical-know-how among most users, for instance to extract and disseminate complex information produced by the system and limited technical support to resolve network issues. Secondly, the results of the PCA models confirm that the strength of VS as a decision-making BI tool can be ascribed to its ability to perform unstructured tasks, collate quality information, enhance total productivity, whereas the operational capacity and functionality of the system is hampered by the System’s inflexibility to be integrated with other IT systems, inability to facilitate new/non-routine/unstructured tasks. Finally, the results of the TTF-based PCA model show that the VS is mostly (in) directly influenced by its operational capacity and functionality features, and its shortcomings is attributable to the System’s incompatibility to meet user’s task profile and inflexibility to execute new task demanded. Based on these findings, the efficiency gain derived by the main users in the focal provincial departments is relatively low vis-à-vis the high cost of implementing the system. The functionality features of the existing VS technology need to be upgraded to allow, for instance, easy accessibility of reports/information with short turn-around time, performance of (un-)structured and non-routine tasks that meet the users’ task profiles. Despite significant inferences produced by the two-pronged quantitative analysis by making use of a small sample, however, the robustness of the inferred results and statistical power of the structural framework will significantly improve by using a larger sample size.