Abstract
Currently, the exploitation of electronic evidence is the highest method that fraudsters use to conceal evidence. This is because of the increase in complexity of digital environments and that the growing reliance on electronic data has introduced new challenges in fraud detection, rendering traditional audit methods inadequate. The rapid advancements in technology, data, and telecommunications integration and usage of business systems has resulted in sophisticated economic crimes (EC) such as fraud. The forensic audit process has become an imperative technique used to detect fraud. Hence, the present study focused on application of data analytics by forensic auditors in fraud detection.
In this study, the researcher followed a qualitative research methodology, using a single case study approach within a South African petrochemical company to achieve the objective of the study. Data relating to 23 forensic data analytics (FDA) engagements were collected between 2024 and 2025. Data were collected and analysed through document analysis, which included forensic audit reports, working papers, policies, and procedures. The study deconstructed the FDA lifecycle into five core phases, namely planning, data identification and acquisition, data cleaning, execution, and reporting. Thematic analysis was used to identify patterns in FDA tools, techniques, and processes. Key results of the study revealed that forensic auditors preferred to gain direct access to data, minimal reliance on external technical assistance, and the use of advanced data analytics tools such as Arbutus, Power BI, SQL Server, and Excel.
The results of the study showed that analytical techniques such as anomaly detection, time-series analysis, and clustering were commonly employed to detect irregularities. Proactive FDA engagements were more prevalent and served as an early detection mechanism, while reactive engagements were critical in substantiating allegations. The research underscores the indispensable role of FDA in modern fraud detection, offering a practical approach to data analytics tools, application, effective techniques, and standardised methodologies to detect fraud. Moreover, the study further highlights the importance of structured methodologies, data integrity, and the integration of advanced data analytics in enhancing fraud detection capabilities.