Companies invest heavily in artificial intelligence but struggle to achieve substantial returns, and the root cause often lies not in the technology itself, but in a lack of leadership and lagging process transformation. Neil Dhar, Senior Vice President of IBM Consulting for the Americas, recently pointed out in an interview with CFO Dive that many organizations face a common dilemma: too many siloed pilot projects and insufficient focus on transforming core business processes.

Currently, companies are under pressure to shift from AI experiments to measurable outcomes. Dhar believes that this shift requires organizations to adopt a more disciplined approach to AI advancement—redesigning legacy business processes before deploying AI, and ensuring tangible results through strong leadership.

"Otherwise, you're just conducting a science experiment, and you won't get very far," he said.

The following is an edited transcript of the Q&A between Dhar and CFO Dive reporter Alexei Alexis, edited for clarity and brevity.

How have CFOs' AI investment priorities evolved?

CFO Dive: How have CFOs' priorities in AI investment changed in recent years?

Neil Dhar: If you look back at the AI wave, it only truly took center stage a few years ago. The initial focus was clearly on education, because most people knew nothing about AI, right?

In the early stages, there was a lot of FOMO (fear of missing out), and companies launched siloed proof-of-concept projects, but these often fizzled out with minimal return on investment. In many cases, companies accumulated considerable costs while chasing the AI dream.

However, I think over the past six to nine months, companies have become more focused on return on investment, and CFOs are at the core of driving this shift.

How does measuring AI ROI differ from traditional investments?

CFO Dive: Compared to traditional corporate investments, does AI present new ROI challenges for finance executives?

Neil Dhar: I believe AI ROI is absolutely trackable, but you must have discipline, just as you need discipline in any past transformation. I think those disciplined companies will be able to capture value, communicate value, and reinvest in value.

Take IBM's own AI journey as an example. We identified business processes that are critical to the company, where there are opportunities to drive real AI benefits. We strive to measure those benefits while also paying attention to customer and employee satisfaction.

Which areas of AI ROI are easier to measure?

CFO Dive: In terms of AI ROI, are some areas easier to measure than others?

Neil Dhar: I think in terms of productivity, some things are relatively easy to measure. For example, if a task originally took X amount of time, can AI help you complete it faster?

Measuring growth and innovation, on the other hand, is always harder. That's a fact.

We recently released aCEO surveywith some very interesting data points. 80% of executives expect AI to drive significant revenue growth by 2030, but only 24% know where that revenue will come from. So, simply put, revenue is harder to measure.

What is the biggest mistake in failed corporate AI investments?

CFO Dive: Some surveys show that many companies stillhaven't seen meaningful returns from AI investments. What is the biggest mistake holding them back?

Neil Dhar: First, I think it's because senior leaders don't truly have a technology-first mindset; AI is delegated to employees three levels down, and the executive team isn't personally involved. Second, it's the approach of "jumping straight into AI projects" without first thoroughly deconstructing business processes. I think these are the two most common mistakes.

You have to be willing to look at your processes end-to-end and ensure there's strong leadership to truly drive tangible results. Otherwise, you're just conducting a science experiment, and you won't get very far.

What role should CFOs play in ensuring AI investment returns?

CFO Dive: What specific role should CFOs play in ensuring AI investments generate real returns?

Neil Dhar: My view is simple: the CFO is the steward of the company's assets. As a CFO, for every dollar of capital you spend, you should demonstrate a return, right? In my view, that return should be around 2.5 to 3 times.

AI is no exception. CFOs must truly focus on ensuring the AI journey produces concrete, tangible benefits.

In many ways, CFOs are responsible for establishing processes to track whether any change is being made. At the same time, CFOs are also responsible for ensuring that any information communicated to the board and investors is backed by auditable data.