AI Agenda Live: EY Consulting on Bringing AI ROI to your P&L

The capabilities AI can bring to companies seem limitless. So do AI token bills, which have skyrocketed, pressuring CFOs and boards to show that such spending can actually boost revenue or profits.

Dan Diasio, EY Global AI Leader and Consulting CTO, has led teams advising the world’s largest companies on their AI transformations. The Information Editor-in-Chief Jessica Lessin sat down with Diasio to discuss why productivity gains from AI aren’t automatically showing up as financial returns and what has to change for them to do so.

From enthusiasm to ROI

According to Diasio, rising token costs haven’t dampened enterprise enthusiasm for AI. But enthusiasm alone isn’t enough to move profit margins. To underscore this point, he referred to two surveys suggesting that many organizations believe AI is paying off, but few can yet trace those returns in their financial results. The first, conducted by The Information, found that a third of organizations reported getting a return from AI that was a multiple of what they spend on it. The second, an EY survey conducted as part of the EY CEO Imperative Series, found that only 11% of CEOs say their companies link AI impact to financial reporting and review it regularly at the senior management level.

“There’s value in increasing productivity, and there’s the value organizations need to be able to prove in their [profit and loss statement], and those are two different things,” Diasio said. “Many organizations I talk to would say AI is significantly increasing productivity. But when the CFO looks at the investment, that productivity needs to translate into two things: Are we making more revenue, and are we taking costs out of the organization?”

Thinking beyond use cases

What this means is that business leaders have to devise more concrete plans for AI  than they did in past cycles.

“I’m seeing this big pivot from ‘let’s go identify some use cases’ to using AI to drive business transformation,” he said.

To illustrate the point, Diasio described how a consumer products company could use AI to cut the time from product idea to market from 18 months to three, a move that would drive revenue by bringing more products to market faster.

“That’s the sort of change that enterprises need to see to be all in,” he said.

The agent handoff

As part of that transformation, organizations are rethinking the relationship between human employees and so-called AI agents and, in many cases, finding unique ways to use agents to create value. Diasio pointed to banks, which have started using agents to assist financial advisors and streamline back-office work. The real value, he said, is in how those AI-assisted touch points build the bank’s relationships with clients and hopefully pay off by  helping it generate more revenue.

“They’re using a lot of these AI capabilities not as a way of reducing costs, but of building brand,” he said of the banks.

Starting fresh

Ultimately, making AI’s payoff visible in financial results is going to require organizations to think of AI less as a technology rollout and more as a driver of measurable results. That means redesigning whole business functions from the ground up, not just automating pieces of them.

“Fast-forward 12 months and companies are going to start thinking about new ways to run their business in parts,” he said. “That means doing the hard thing: addressing the processes that need to be entirely rebuilt.”

That’s a harder path than simply adopting new tools, he said, but it’s the one that actually shows up on the P&L.