At a Glance

  • According to a study released by SAP on Wednesday, despite the expanding scale of AI applications in enterprises, most companies still struggle to define and measure return on investment (ROI). The study, conducted by SAP in collaboration with Oxford Economics, surveyed 2,600 directors and C-level executives across mid-sized and large enterprises in 13 countries.
  • The survey shows that although satisfaction with AI investment returns has risen a few percentage points year over year to 69%, nearly the same number of respondents said they are not confident that the technology they have deployed is reaching its full potential.
  • Sean Kask, SAP's Chief AI Strategy Officer, told CIO Dive that companies are in a bind because ROI itself is difficult to quantify. Companies may assess whether a project's feasibility matches their ROI matrix. Kask suggested: "Look at the use cases you have, evaluate the value you think you can derive from them, and weigh that against the effort required to implement them."

Deep Insights

As companies become more mature in their experience with AI technology,confidence in AI is growing, but different members within organizations stilldisagree on how to define success or measure its business value.

Although SAP found that satisfaction with AI value is rising, the ERP vendor also found that cost is not one of the primary benefits companies are achieving. Respondents reported that AI applications are less effective in driving cost efficiency or productivity than in helping employees gain insights, make decisions, or interact with customers.

Kask said: "Cost is not the primary benefit driver of AI, but it is certainly part of it."

Compared to last year, AI usage has increased slightly, with an average of 30% of tasks now completed with AI assistance, up from 25% last year. The survey projects this will rise to 48% over the next two years. Only 18% of companies said they are using end-to-end, cross-functional AI deployments rather than limiting it to single-task scenarios.

SAP and Oxford Economics found that U.S. companies will spend an average of $37.2 million on AI this year and expect to increase spending by 46% over the next two years. U.S. companies reported an average AI investment return of $9.9 million this year, projected to reach $26.5 million over the next two years.

Earlier this month,a KPMG surveyfound that in the coming years, leaders at mature enterprises will focus more on scaling AI and defining its role in the business. But Kask said data, skills, and governance will remain the biggest challenges to achieving ROI.

Kask noted: "As companies begin to scale and gain experience, especially when using agents, they will discover many shadow agents or shadow IT agents that may have accessed data they should not have, or performed system actions on behalf of users that may not be auditable."

The deployment process can expose weak points that companies need to focus on, or reveal areas where best practices should be continuously reviewed. The study found that AI is fundamentally reshaping organizations that invest heavily in the technology. Kask believes that companies that succeed are those with board-level AI literacy and a willingness to restructure their systems.

Kask said: "The key is the allocation of scarce resources—not throwing everything into AI, but doing it in a way that maximizes ROI. I think this needs to be done at the board level, and frankly, it is about selecting the areas that can truly change processes in a strategic way, and then measuring the results."