metamorworks - stock.adobe.com
Deriving ROI from AI is more than just monetary value
While some enterprises can see the business value from their AI initiatives relatively quickly in terms of money value, others see it in the efficiency of their employees.
NEW YORK -- In October 2024, enterprise health platform vendor Oura launched the Oura Ring 4. With popularity of the ring came a significant problem: the company had more demand for ring support than customer service agents who could deliver it.
The problem of not having enough support led Oura to turn to conversational AI platform provider Decagon to launch a new chat system to help support agents deal with the flood of emails and support interactions they were bombarded by.
"We had to do this in about four weeks," said Tom Hale, Oura CEO, during a presentation about AI and ROI at the Reuters Momentum AI New York 2025 conference on Monday. "We were able to keep our customers happy." He added that customer service agents could now service tickets despite a dramatic spike in demand for support from October to December.
"We were able to do this without hiring 500 additional support folks because we literally couldn't hire them and train them in the time frame," Hale said. "So, the savings here will amount to about $15 million."
The ROI problem
Hale and his team used Decagon's technology to discover a possible ROI on AI quickly.
Since the advent of widely available generative AI tools over the last few years, many observers have inferred that many enterprises have yet to find ROI using the technology.
But for Hale and many enterprises, ROI depends on the application and how ROI is defined.
"Sometimes when you think about the generic productivity use cases, it's tough to see the ROI of one person using AI once a day," Hale said in an interview with Informa TechTarget.
He added that in Oura's case, "necessity was the mother of invention."
"We are going to either disappoint our customer or we are going to succeed," he continued. "That urgency is a big reason why we knew what we had to do. We saw the path of success and were able to move quickly to get to it."
When discussing ROI, it's also essential to consider the industry, said Cristina Mancini, CEO of Black Girls Code, in an interview with Informa TechTarget.
"This is a unique time where technology is affecting every industry and every industry has its ROI," Mancini said. "That's a question you must ask yourself: what am I trying to solve for, and am I making something useful and safe?"
ROI and efficiency
ROI is also not always measured in terms of money.
For example, in the healthcare industry, a nurse practitioner or a doctor could be using AI to better perform the parts of their job that might require making calls to remind patients to come to an appointment or following up after an appointment. Such tasks might only be 12% of the nurse or doctor's job.
"The challenge of AI is it doesn't replace entire people. It replaces fragments of what we do," said Daniel Barchi, chief information officer at CommonSpirit Health, a nonprofit Catholic healthcare provider organization, in an interview with Informa TechTarget. "Mostly the fragments that are behind a desk or a phone."
He added that using AI technology to replace tasks for nurses and physicians, such as calling patients by phone, improves quality, allowing the nurses to focus more on direct care.
"Our patients have a better outcome when the nurses spend all of their time on direct patient care, but it challenges the ROI because it's an increased cost of the technology, but we are not replacing 12% of a person,” Barchi said.
Moreover, it's important not to downplay the ROI of efficiency when AI technology is used to perform a job better, Sydney Klein, senior vice president, chief information security officer, and head of IT Core Services at Bristol Myers Squibb, told Informa TechTarget.
"If you get people to figure out that AI is their superpower and they have figured out how to be much more effective because of AI, and it is something that makes them more efficient, then they're going to be looking at everything they do with a different perspective," Klein said. "ROI is going to be hard to tap into directly to prove that correlation, but I think that is going to have an applicable effect on the big business problem you can solve with AI."
Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.
