Data Analytics projects have become increasingly critical within the field of Supply Chain Management. Deriving data-driven decisions is pivotal not only for optimizing supply chains and ensuring operational efficiency but also for realizing substantial cost savings, identifying potential risks at an early stage, and implementing proactive measures. These strategic actions significantly enhance the financial performance and profitability of organizations. Given these advantages, there is a pronounced need for research to investigate how Data Analytics is practically applied within companies, and whether established frameworks such as CRISP-DM, SEMMA and KDD from the literature are indeed utilized in practice, and if so, how and why they might be adapted. There often is a discrepancy between the theoretical constructs of these models and their practical implementation. This study commences with an introduction that thoroughly outlines the problem statement, objectives, research questions and methodology. It is followed by a theoretical chapter that conducts an extensive literature review on the key topics of Data Analytics, Supply Chain Management and Project Management. This review is aimed at defining and deepening the understanding of these concepts while exploring their applicability. In the empirical portion of the study, focus group discussions with industry experts are employed to obtain in-depth insights into the practical execution of Data Analytics projects. This section examines how these projects are carried out in real-world scenarios, the extent to which theoretical frameworks are applied and which elements are adopted or modified. The goal is to juxtapose these empirical findings with the theoretical models, identifying both congruencies and deviations. The results of the study demonstrate that while the phases of the theoretical models are often applied in practice, there are significant adaptations required to meet the unique needs and conditions of organizations. The CRISP-DM model, which inherently includes iterative processes, is frequently implemented in its foundational form. However, in practice companies tend to adopt a flexible, iterative approach rather than adhering strictly to predetermined processes, ensuring that requirements are accurately understood and effectively executed. Although the core phases of the three models exhibit substantial similarities, differences emerge in their focus – particularly between businessoriented and technically-focused approaches. Overall, the CRISP-DM model most closely aligns with the methods employed in practice, whereas SEMMA and KDD, with their strong technical emphasis, are less commonly applied in their original configurations. The study concludes by recommending a closer integration of theory and practice and suggests that future research should engage with larger and more diverse samples, focusing on the specific customization of these models across various industry sectors.
Date of Award | 2024 |
---|
Original language | German (Austria) |
---|
Supervisor | Patrick Brandtner (Supervisor) |
---|
Vorgehensmodelle in Data-Analytics-Projekten im Supply Chain Management - CRISP-DM und Co. in Theorie und Praxis?
Geiblinger, J. (Author). 2024
Student thesis: Master's Thesis