Key Takeaways
Ask most companies what they want from AI, and the answer sounds like someone standing by a track with a stopwatch: faster service, shorter workflows, fewer repetitive tasks, higher productivity. These are useful and relatively easy to measure, and can be a big boon for companies.
But artificial intelligence (AI)‘s promise was never just about helping people work faster. It was about helping them work better. And that involves more than “freeing humans to do what humans do best.” It means giving humans the time for high-value work that creates impact — and true ROI — for the company. This could be everything from innovating on a product to finding the next opportunity hiding in plain sight.
“There’s been an initial push with generative AI and AI to find efficiency in the way that people do their jobs. And efficiency has been a good metric, but it’s not one that turns into true realized value,” said Ben Richards, managing director, Canada customer growth and transformation at Salesforce. “If I can save someone 10 or 15 minutes of time, what are they doing with that time? The most valuable use cases are when we apply AI to very tangible areas of return.”
What does high-value work look like?
High-value work, by definition, creates impact for your company. It fuels innovation. It deepens customer relationships and increases employee retention. It makes your company stand out in the crowd.
To understand what this looks like in real life, let’s look at how a few organizations are seeing ROI from AI.
A bank’s wealth advisors can spend more time with clients
RBC Wealth Management, a division of Canada’s largest bank, was facing a challenge: The company had doubled its business between 2018 and 2025, and wanted to double it again — in half the time. But with more wealth to manage than ever and a shortage of experienced financial advisors, the bank couldn’t meet the growing demand through hiring alone.
Just as challenging, RBC’s 2,200 wealth advisors were already swamped. Between manual customer relationship management (CRM) updates, portfolio research, meeting prep, and note-taking, they didn’t have enough time to have in-depth strategic conversations with customers, let alone take on new clients. And their work was slowed by disconnected data and apps.
The company realized its advisors needed more tools. So, RBC deployed Agentforce, Salesforce’s platform for building and deploying AI agents. It created an agent that preps advisors for meetings, creating one-pagers complete with portfolio details, upcoming tasks, and even personal information like the client’s favorite restaurant or upcoming anniversary.
What used to take an hour of digging through client data, now takes less than a minute — and frees advisors to spend more time with each client. “The advisors can posit better strategies to their clients, and they can potentially have an hour-long meeting, instead of a half hour, because they don’t have so much packed in their calendar,” said Richards. It’s high-value work that is helping RBC’s wealth management division grow.
A medical center can focus on better patient outcomes
Meanwhile, Sarah Duvall, a nurse practitioner at the University of Rochester Medicine (URM), uses AI to help with some of the highest-value work of all: improving patient outcomes.
Duvall, who’s worked at URM for 25 years, recently joined the surgical oncology team, and one of her first assignments was to look at post-surgery readmission rates, which were high.
She took a class on AI for healthcare professionals at the University of Rochester’s Simon Business School, and created an AI tool to analyze research and data. She especially wanted to know whether a prehabilitation program — which prepares patients for surgery by doing things such as eating better and getting gentle exercise — would help.
Using a variety of AI tools, Duvall compared the cost of readmissions (upwards of $3,000 a day) to that of a prehabilitation program (about $2,000 per patient, total). “‘When I crunched the numbers using AI, I could see the correlations and operational bottlenecks quickly. I didn’t have to dig through pages and pages of data,” Duvall said.
Once Duvall had the data, she used AI to create a proposal in language that business leaders could understand. It was work that, as a busy healthcare professional, she wouldn’t normally have had time to do.
The result? Her proposal showed that if URM implemented a prehabilitation program, it could save at least $200,000 per year and more than $900,000 over three years. It could also save patients — and their families — a lot of agony. Her team is piloting the program this summer and hopes to secure a grant to fund the program soon.
The best way to measure the ROI of high-value work
High-value work often involves judgement, creativity, intuition, and relationship-building — things that, by their very nature, are hard to measure.
It’s true that RBC Wealth Management will show value by doubling its business. And Duvall has shown it in dollars and cents. But measuring the worth of deep human work, said Dan Keating, clinical associate professor of information systems and AI at the Simon Business School, can be tricky for a simple reason. “We are not great at articulating the value we bring as humans; we’re just not very good at saying why our work is better than something AI has produced,” he said.
Richards said it’s probably not the work itself you want to measure. It’s the outcomes. He pointed to the example of a telecommunications company. One of the most important metrics for such a company is customer attrition. But if AI agents handle basic customer requests — quickly — and free human service reps to handle more complex inquiries, the company could boost customer satisfaction scores and reduce attrition. “You can target the high-level outcomes you want to achieve, and work back from that,” said Richards.
So, go ahead and measure speed and efficiency if you like. But true payoff lies in the impact.
Companies need to redesign work for ROI
The promise of AI has always been that it will free people to do more high-value work. But don’t assume this will just happen on its own.
In a recent Harvard Business Review article, researchers found that AI tools didn’t reduce work. They intensified it. Employees who used AI worked at a faster pace, took on more work, and worked longer hours. The upshot? They also experienced workload creep, cognitive fatigue, burnout, and weakened decision-making. In other words, the exact opposite of high-value work.
If companies want to reap the true benefits of AI, they need to design for it. They need to make sure the newfound time is meaningfully deployed. That can happen in several different ways.
One of the first things companies need to do, Richards said, is revisit org charts. “What agentic AI is giving us is the ability to completely rethink the way an org chart needs to be built,” he said. “If we’re going to take the low-cognitive tasks off people’s plates and shift them to higher cognitive tasks, we also have to shift the way the org structure is built.”
That could mean managers oversee both AI and people at the same time. And frontline workers could actually become managers of AI agents. The idea: to free up humans by letting agents handle more of the busy work.
That means protecting time for high-value work
Companies then need to protect the newfound time to make sure humans use it well. The HBR article suggests ways to do this. Businesses can build intentional pauses into workflows. Before a major decision is made, for example, employees could be required to step back and examine other viewpoints. Or companies could protect “focus time” by batching non-urgent notifications and holding them until natural breakpoints.
But it could also mean giving humans time to collaborate and work together — an idea that Nikita Skylarov mentioned in a comment on a LinkedIn post. The cofounder of the software development company MokuHub, Skylarov wrote that in his company, “the time [AI] freed up didn’t go to working alone. It went to working together on harder problems.”
Even if surplus minutes or hours are spent alone, companies can protect this time for employees to experiment, coach, plan, or let their imaginations run free. As Keating said: “The qualities that make us human are what have allowed incredible innovation throughout history. Every once in a while, we make these massive leaps of creativity that AI can’t make. AI can mimic the works of Shakespeare. It never could have written the works of Shakespeare.”
Richards pointed out that many companies employ highly educated people in knowledge roles, and may not be taking advantage of all their expertise. “If we can open the door for critical thinking and the cognitive load that we’ve almost been pushing away,” he said, “that’s where high-value work can really come out.”
One of the biggest benefits of high-value work
When employees use their knowledge and skills at work, they’re more likely to feel valued and stay at their organization. That reduces turnover, which saves companies money.
But when employees are happy, they also do their best work. Take Duvall, the nurse practitioner, as an example. AI not only saved her time but also amplified her skills of synthesizing research and communicating well. “And that’s allowed me to do things that are more joyful in my work,” she said.
Chiefly: spending time with patients and creating better outcomes. It’s hard to imagine a better ROI than that.
FAQs
High-value work is work that people can do when they’re freed from mundane tasks by AI. It involves human judgment, creativity, and relationship-building, and creates real impact for the company.
Focus on high-level outcomes instead of just tracking speed. The biggest ROI comes when your team can do meaningful work to foster retention, customer satisfaction, innovation, or growth.
Humans often struggle to explain why their work is better than AI’s output. That’s because human work involves intuition, wisdom, and emotion, which numbers can’t easily track.
When AI frees up time for people, employees sometimes just fill it with more busywork, or their workload intensifies so much they’re at risk of burnout. But companies can intentionally design roles to give employees time for high-value work.
Companies can build intentional pauses into daily workflows. They can also batch non-urgent notifications, and hold them until natural breakpoints, to protect an employee’s time.
