A Review on Understanding Data Governance Failures in Analytics Systems: Insights from Expert Interviews and Root-Cause Thematic Coding

Abstract

This study investigates the persistent problem of weak analytics outcomes in organizations despite substantial investment in data platforms, dashboards, and business intelligence tools, arguing that the deeper source of breakdown often lies in data governance failure rather than in technology alone. The purpose of the research was to examine how specific governance failures affect analytics system effectiveness and to identify which failure dimensions are most damaging in organizational settings. Using a quantitative, cross-sectional, case-based design, the study collected survey data from cloud and enterprise analytics cases through 250 distributed questionnaires, of which 214 usable responses were analyzed, yielding an effective response rate of 85.6%. The sample consisted of professionals directly involved in analytics and data processes, including Data/BI analysts, IT and systems staff, compliance officers, and reporting managers. The key independent variables were Data Quality Failure, Accountability and Ownership Failure, Policy and Compliance Failure, Access-Control and Security Failure, and Metadata and Documentation Failure, while the dependent variable was Analytics System Effectiveness. The analysis plan combined descriptive statistics, Cronbach’s alpha reliability testing, Pearson correlation analysis, and multiple regression modeling in SPSS. The findings showed high perceived exposure across all governance failure dimensions, with Data Quality Failure recording the highest mean score (M = 4.08, SD = 0.71), followed by Metadata and Documentation Failure (M = 4.02, SD = 0.69), while Analytics System Effectiveness remained comparatively low (M = 2.64, SD = 0.83). Reliability was strong across constructs, with Cronbach’s alpha ranging from 0.82 to 0.91 and an overall instrument alpha of 0.93. Correlation results revealed significant negative relationships between governance failures and analytics effectiveness, led by Data Quality Failure (r = -0.68, p < .01) and Metadata and Documentation Failure (r = -0.64, p < .01). The regression model was statistically significant, F (5,208) = 46.37, p < .001, explaining 52.7% of the variance in analytics system effectiveness (Adjusted R² = 0.527). Data Quality Failure emerged as the strongest predictor (beta = -0.31, p < .001), followed by Metadata and Documentation Failure (beta = -0.27, p < .001). The study implies that organizations seeking more trustworthy and decision-supportive analytics should prioritize governance reforms in data quality, documentation, accountability, and policy enforcement rather than relying only on technical upgrades.