This study identifies the client and machinery-related factors and examines their interrelationships in construction procurement, employing a systems-thinking approach. Data from stakeholders in the KwaZulu-Natal construction industry were analysed through statistical methods and Applied Systems Analysis (ASA), with causal loop diagrams (CLDs) illustrating how procurement performance evolves via reinforcing and balancing feedback loops. Findings indicate that clients’ emphasis on cost and timely delivery strongly drives procurement decisions but can weaken decision-making, reduce communication, and increase project variations. Machinery-related challenges, including poor maintenance, high hiring costs, and limited access, further exacerbate inefficiencies. When considered together, balancing loops emerge, where strong governance, effective communication, preventive maintenance, and operator training stabilise and improve outcomes. The study contributes theoretically by advancing a systems-based perspective that integrates client governance and machinery management, and practically by identifying leverage points for policymakers, clients, and contractors, such as improving governance, embedding maintenance practices, aligning priorities with resource planning, and institutionalising feedback. Procurement effectiveness is thus shaped by interdependent governance, decision-making, resource management, and systemic learning. Keywords: Applied Systems Analysis; Client; Construction; Machinery, Procurement; Governance

This study develops an empirically derived, context-specific taxonomy of risks associated with green building projects in Lagos State, Nigeria, and explains how these risks are structurally organised in an institutionally underdeveloped market. A quantitative cross-sectional survey yielded 152 valid responses from registered architects, quantity surveyors, builders, engineers and project managers (72.4% response rate). Twenty-five risk variables identified through a systematic literature review were rated on a five-point scale and analysed using descriptive statistics and principal component analysis (PCA) with varimax rotation, applied as a dimension-reduction procedure, with the number of components to retain determined by Horn’s parallel analysis. Findings were interpreted using the Technology-Organisation-Environment (TOE) framework and systems theory. Financial risk attracted the strongest agreement as a relevant risk (M = 4.14), followed by design and construction risks (M = 3.80). Parallel analysis supported retaining three components, which together explain 65.37% of the total variance: operational, performance and lifecycle risks; technical and external interface risks; and design, regulatory, market and financial delivery risks. Although financial risk attracted the strongest agreement of any single item, it did not constitute a distinct risk dimension; it was subsumed within the broader front-end delivery component, whereas operational and lifecycle risks accounted for the largest share of common variance. The taxonomy therefore reveals a divergence between the strength of agreement a risk attracts and the share of common variance it accounts for. Findings support lifecycle costing, incentive-based regulation, post-occupancy performance management, and green building risk management within continuing professional development. Convenience sampling, a Lagos-only sampling frame and an early-career respondent bias limit generalisability. This study presents an empirically derived parallel-analysis-verified taxonomy of green building risks, advancing TOE- and systems theory-based understanding of green building risk in developing economies.