Jeevan ThankappanNearly one-third of artificial intelligence (AI) projects may be scrapped by 2025 as costs outpace perceived value, according to recent findings from Gartner. This challenge has led many CEOs to rely on CFOs to determine whether AI initiatives are worth the investment. “Determining the return on investment (ROI) of AI can often feel like solving a puzzle with missing pieces,” says João Carvalho, Managing Director of SAP Concur in Southern Europe (Spain, Portugal, Greece, and Turkey), the Middle East, and Africa.
“AI in the workplace is still in an early phase, with few established templates for implementation and success,” he adds. “While this presents exciting new possibilities, it also complicates how CFOs can define and measure ROI, given factors like variable market conditions, adoption rates, and high operational costs.”
For finance leaders facing these hurdles, Carvalho offers four tips to help organisations realise AI’s potential value:
- Adopt multi-disciplinary standards for success: Achieving strong returns on AI investments requires maintaining high standards across multiple domains, akin to an Olympic decathlete’s multi-event approach. Deloitte reports that top-performing companies excel in data management, results tracking, security, privacy, and governance. High-quality data fuels AI models, consistent results tracking informs adaptable strategies, and robust security measures ensure data protection and ethical standards.
- Secure early wins and leverage initial insights: With AI’s potential business impact, it’s easy to focus on the big picture while overlooking process details. Focusing on early wins can build foundational evidence in an AI journey, justifying continued investment and keeping stakeholders engaged. Expanding successful pilot projects quickly demonstrates value. However, calculating AI ROI remains more art than science, requiring CFOs to balance financial and strategic benefits. Early analysis also enables leaders to align AI insights with broader business goals, helping to justify long-term costs.
- Tailor measurement metrics to organisational goals: AI benefits vary, so standardised ROI measurements may not always apply. Some may not appear on traditional spreadsheets or in key performance indicators (KPIs), making ROI assessments complex for finance teams. Beyond core financial metrics, factors like quality improvements, innovation, and compliance should be considered. A report by IDC reveals that it typically takes organisations around 14 months post-deployment to achieve AI ROI, underscoring the importance of patience and tailored metrics.
- Acknowledge the human aspect of AI ROI: Defining AI ROI can be challenging, particularly with talent gaps. With 52% of companies citing a lack of skilled workers, AI projects demand investment in both technology and workforce training. However, AI adoption can help reduce training costs and attract young professionals eager to work with advanced tools. According to the Institute of Management Accountants, overlooking AI can risk losing young finance talent. The aim is to enhance team capabilities and address talent gaps, not replace human roles.
“Evaluating AI projects with a portfolio-based approach—rather than assessing one project at a time—allows organisations to consider factors like talent attraction and development,” Carvalho explains. “Collaboration across departments, especially with IT, is essential.”
According to the SAP Concur CFO Insights report, 53% of finance leaders agree they need to partner with IT heads on AI, 49% highlight collaboration on digital transformation, and 24% on cybersecurity. “Organisations today have many options for AI implementation, and a key consideration is whether to build or buy,” he notes. “Developing custom software offers exclusivity but requires a committed roadmap, while buying can bring quicker results.”
“Partnering with experts to understand the ROI equation can help you calculate the best mix for your organisation—wherever you are on the AI journey,” Carvalho concludes.
