From accounting systems to business performance: A serial mediation model with IPMA evidence from Vietnamese SMEs

This study examines how organisational and technological antecedents shape management accounting information quality and adoption level among Vietnamese small and medium-sized enterprises (SMEs), and how these characteristics influence decision quality and business performance, with business environment uncertainty as a moderator. Using a cross-sectional survey of 503 Vietnamese SME managers, the study applies partial least squares structural equation modelling and importance-performance map analysis to test a five-antecedent serial mediation model grounded in contingency theory and information processing theory. All ten hypotheses were supported. Information technology capability was the strongest driver of accounting information quality (β = 0.508), and market competition intensity was the strongest driver of adoption level (β = 0.385). Decision quality was the strongest direct driver of business performance (β = 0.558), and business environment uncertainty positively moderated this relationship (β = 0.238). All five serial indirect paths to business performance were statistically significant. The importance-performance analysis identified information technology capability as the highest-priority investment area, combining high importance with low realised performance (52.01 out of 100). To our knowledge, this study offers one of the first comprehensive tests of a serial mediation model linking these antecedents to business performance through accounting characteristics and decision quality in Vietnamese SMEs, one of the first importance-performance prioritisation maps for managers in this context, and extends contingency theory by identifying business environment uncertainty as a pure moderator rather than a direct antecedent. Findings suggest Vietnamese SME managers should prioritise digital accounting infrastructure investment ahead of expanding accounting tool adoption, and that government support for digital accounting tools may be warranted. The cross-sectional design limits causal inference, partly addressed through theoretically grounded antecedent ordering, predictive validity testing, and multi-method bias assessment; the single-country scope reflects Vietnam's structural representativeness of wider ASEAN transition economies.