A Longitudinal Study of Technical Debt Accumulation and Management in Nigerian Commercial Software Products

Technical debt is a central construct in software engineering, yet longitudinal empirical studies of technical debt accumulation and management trajectories in commercially deployed software products remain rare globally and entirely absent from the Nigerian software engineering literature. This dissertation conducts a longitudinal study of technical debt accumulation and management in Nigerian commercial software products, making original empirical and theoretical contributions to the technical debt research stream. A 24-month longitudinal study tracks technical debt metrics across 18 commercial software products maintained by 12 Nigerian companies in Lagos (South West), Enugu (South East), and Abuja (North Central), covering fintech, edtech, and enterprise software domains. Technical debt is measured using automated static analysis (SonarQube metrics: code smells, security hotspots, duplications, test coverage deficit) combined with qualitative developer interviews conducted at six-month intervals. The study constructs individual debt accumulation trajectories and applies trajectory clustering to identify four debt growth archetypes: Managed Growth, Debt Explosion, Periodic Reset, and Structural Stagnation. Regression analysis identifies delivery velocity pressure (beta = 0.44), team turnover rate (beta = 0.37), and absence of refactoring budget (beta = 0.31) as the strongest predictors of rapid debt accumulation. The dissertation proposes an original Technical Debt Governance Cycle theory, describing the conditions under which organisations transition between debt management and debt escalation states. Keywords: technical debt, longitudinal study, software products, Nigeria, debt management