A spectral analysis extension to DEMATEL for strategic leverage points identification

Abstract Efforts to intervene in complex systems often emphasize influential factors, yet system behavior is equally shaped by the relationships among them. Methods such as Decision‐Making Trial and Evaluation Laboratory (DEMATEL) map causal structures but remain descriptive and do not identify which relationships provide the greatest leverage for change. This study develops a spectral sensitivity framework that extends DEMATEL from mapping to prescriptive analysis. We link system amplification to the maximum eigenvalue of the total relation matrix and derive its sensitivity to each relationship, classifying modifications as amplifying responsiveness or enhancing stability. Our contributions are threefold: (i) a mathematical framework for relationship‐level leverage point identification with closed‐form sensitivity results; (ii) a demonstration that its assumptions hold in practice; and (iii) an open web application that automates the analysis and delivers actionable outputs. A case study on challenges in adopting energy management systems demonstrates how the method identifies critical intervention pathways for system change.