How do you know your training dollars are supporting business outcomes?

In today's rapidly evolving workplace, upskilling and reskilling are more important than ever. At a global level, companies spend more than $350 billion a year on learning and development. Today, more than one in five L&D professionals say their number one priority is training employees how to work alongside generative AI. 

Besides getting up to speed on gen AI, workers will also have to take on new responsibilities and develop new skills as parts of their jobs are automated. This dramatic paradigm shift has made L&D the most important function within every organization, argues Matt Beane, an assistant professor at the University of California at Santa Barbara, and author of The Skill Code: How to Save Human Ability in an Age of Intelligent Machines.

"Without top-notch L&D, introducing gen AI into work is a bit like yelling fire in a crowded theater: It will tear organizations, teams, and careers apart from the inside," Beane writes. "CEOs need to quadruple down on L&D as their new, #1 most critical business partner, and hold them to a higher standard than ever."

The increased importance of training puts greater pressure on HR professionals to demonstrate a tangible return on their L&D investments. They can no longer measure the impact of internal education programs using statistics centered around attendance or other traditional metrics, says Anthony Onesto, chief people officer for consumer insights firm Suzy. Organizations need better metrics that are more closely tied to business results.

"HR often falls into a trap of pulling together metrics like how many people attended [a training session]," he says. "That doesn't give you any measure of effectiveness. To me, it's the business results. And that always ties to productivity and profitability." 

Bottom line results matter most

For L&D investments to show their worth, the impact needs to be measurable in ways that adds real value to the business, notes Paul Daugherty, chief technology and innovation officer for Accenture, and co-author of Human + Machine: Reimagining Work in the Age of AI. But some types of impacts are more easily measured than others.

For example, developers trained on how to use GitHub Copilot completed coding tasks in 55% less time and with fewer errors than those using standard coding practices. The ability to produce fast, bug-free software allows devs to take on more projects, often with positive impacts on a company's bottom line. 

But simply letting your devs loose on Copilot won't buy you much, Daugherty warns. 

"Unless you show people how to use it, they're not going to get much more productive," he says. "You need the right data and training to make it happen."

Another easily measurable outcome is sales. Dextego's AI-based sales training software – which provides instant feedback on a representative's knowledge, techniques, and persuasiveness – allows its customers to close deals 30 percent faster on average, notes company cofounder Ioanna Onasi. Shortening sales cycles allows you to close more deals and bring in more revenue, she adds.

"Most of our customers are more focused on quick onboarding and ramp-up than on retention," says Onasi. "It's all about how fast the training allows them to achieve their goals in the first 30 to 90 days."

Daughtery says Accenture helped a major telecommunications firm implement generative AI in its customer contact centers. The technology made it easier for support agents to understand why customers were calling, their history of interactions, and the types of equipment they had on hand. That in turn enabled agents to find the right solutions faster. 

The result was easily measured: a 30 percent increase in the number of calls handled, accompanied by a 60 percent increase in customer satisfaction. 

It all comes down to developing the right skills to enable people to use these technologies to their full advantage, Daughtery says. 

"To succeed in AI, you need to invest more in your people than in technology," he adds. "Once you understand your workforce needs, you can look at the workers themselves and figure out the skills they'll need and the learning platforms you need to put in place for them." 

The time to gather data is now

But many training outcomes aren't so easy to measure. In those cases, L&D pros will need to work closely with business unit managers to determine what metrics matter most to them, says Theresa Fesinstine, founder of peoplepower.ai, a consultancy that helps HR professionals leverage the power of AI. 

"The question I would ask is, if you wanted a meaningful report about the effectiveness of AI usage a year from now, what data would you want to start tracking today?" she says. "Pick three projects you believe will help you gain efficiency, and rate your outcomes every month. Do that for a year, and you'll gain some deep insights."

It's not always easy to draw a direct line between L&D programs and downstream business results, admits Dr. Dieter Veldsman, chief scientist for the Academy to Innovate HR. But forward-thinking organizations are gathering data now, with the hopes of drawing inferences from it down the road.

"People analytics can play a strong role, but you need to build those models using large data sets and improve them over time," he says. "Because we trained you on x or y, something improves, and that leads to something else improving later. Good organizations are taking a stab at it, because having some intelligence is better than just feeling your way in the dark."

Training for the jobs of tomorrow

Many employers have started to take a grassroots approach to measuring AI's impact on productivity, says Alexandra Adams, senior product marketing manager at Guild. 

"Companies are thinking about how they can leverage AI to become more efficient, and free up time for more creative problem solving," says Adams. "One way employers have been able to prove impact has been to create two groups of employees, equip one group with AI skilling experience and the necessary tools, then compare how this group performs on a series of tasks relative to the control group."

Ultimately, the value of AI upskilling to the organization may show up in the form of time savings and improved performance metrics, she adds. 

"When you're thinking about building skilling pathways, you want to ensure people are moving into roles that are more future-resilient, and delivering the right skilling experiences to get them there," she says. "AI will open up huge potential for future roles, it's just a matter of equipping people to thrive in those environments."

About Dan Tynan

Dan is an award-winning journalist who has been writing about technology since long before we all started carrying a computer in our pockets. Along the way he's been an investigative reporter, feature writer, copy editor, columnist, executive editor, and editor-in-chief for several national publications.