Strategic Implementation of IT Governance for Optimising Information Systems Performance

Abstract

This article presents a new reinforcement learning approach to optimise the use of IT governance resources to enhance information system performance. Despite enormous IT infrastructure investments, organisations might fail to recoup the investment in terms of quantifiable performance enhancements due to ineffective governance practices. We developed a Q-learning system that learns dynamic optimal policies of resource allocation over the most significant five IT governance dimensions: Strategic Alignment, Value Delivery, Resource Management, Risk Management, and Performance Measurement. Using ESG data sets of 12 countries with various levels of digital maturity, our system learned governance dimensions yielding maximum return on investment at every level of maturity. Results indicate that the emerging economies achieve optimal performance improvements of as much as 99% through emphasis on Strategic Alignment and Resource Management, while developed economies achieve this through more balanced distributions with greater emphasis on Risk Management. The Pareto frontier analysis confirms that our Q-learning method reaches optimal levels of resource utilisation efficiency. Cluster analysis determines sharp-cut modes of governance by levels of economic development. This research offers an evidence-based adaptive approach to IT governance implementation that takes into consideration organization maturity and context to enable decision-makers to deploy scarce resources in order to realise peak information systems performance.