AI GUIDE FOR BOARDS IN SINGAPORE
Section II: Strategy and Value Creation
5.2 Measuring Return on Investment
5.2.2 Recognising intangible value and warning indicators
5.2.1 Considering the total cost of ownership
AI investments may generate value that is not immediately reected in short-term nancial metrics. Intangible strategic value can derive from: • Improved customer experience. • Enhanced employee experience. • Better risk management. • Increased organisational ability. • Stronger competitive positioning. • Faster and more informed decision- making. While these outcomes may be di cult to quantify directly, they may contribute materially to long-term enterprise resilience and competitiveness. Boards should ensure that management considers broader strategic and intangible bene ts. On the ipside, boards should monitor for recurring issues that frequently undermine AI value realisation. Warning indicators may include: • Extended pilot phases without operational scaling. • Underestimation of data and integration complexity. • Fragmented investment across low-impact initiatives. • Poor adoption despite technical deployment. • Lack of business ownership. • Weak alignment with strategic priorities. ese issues o en indicate weaknesses in prioritisation, governance, organisational readiness and execution.
Boards should require visibility of the full lifecycle cost of AI initiatives, not only initial investment. Pilot initiatives are typically conducted in controlled environments and may not fully reect the operational complexity, usage patterns or support requirements of scaled deployment. monitoring, cyber security controls, workforce enablement and ongoing assurance. ese may include usage-based consumption costs, integration with enterprise systems, data preparation and governance, model Boards should require AI investments to be assessed using a total cost of ownership approach. AI-related costs frequently extend well beyond initial so ware acquisition and may include: • Data acquisition and preparation. • Data governance and compliance. • Infrastructure cost, including compute and storage. • Cloud and token consumption cost. • Cyber security enhancements. • Model monitoring and retraining. • Integration with enterprise systems. • Talent and operating costs. • Ongoing governance and assurance activities. Boards should ensure that business cases distinguish between experimental costs at pilot stage and full costs associated with scaling and ongoing operations. Failure to do so may result in overstated returns and underinvestment in critical capabilities such as integration and change management
46
Made with FlippingBook Digital Proposal Creator