‘Upskilling’ Your Workforce? Start by Measuring the Skills They Have Now
Charles Field, a network planning engineer at AT&T in Alieso Viejo, Calif., has been picking up new skills since his teens, when he taught himself how to code. He joined AT&T as a technician in 1989 and, after a series of promotions and lateral jumps inside the company, he’s now using a combination of A.I. and human insights to develop analytical forecasting tools for his team. Along the way, Field has kept learning, with in-house AT&T courses, certifications, and nanodegrees from outside sources like Coursera and Udacity, and training from LinkedIn Learning on “everything from programming to leadership,” he says. All that studying, Field adds, “has put me exactly where I want to be.”
As technology keeps racing ahead, putting the right people in the right jobs gets tougher—especially since, in this job market, there just isn’t enough skilled talent to go around. That’s why the number of companies striving to train their current employees is climbing from 55% in 2018 to 76% by next year, according to a ManpowerGroup study of 1,050 U.S. enterprises called “Humans Wanted: Robots Need You“.
Great, but there’s a catch: Without a clear idea of the skills people have now that might be a great fit in a different job—and which ones they need to gain or develop—simply putting everyone through the same training is unlikely to work well. Manpower’s research shows that assessing employees’ current capabilities, so each person can get the most relevant and useful training, “increases the likelihood of placing the right person in the right role from 50% to 80%.” Yet fewer than half (48%) of U.S. employees in a recent Manpower poll said their skills have actually been assessed.
Assess and evaluate
Mindful of that gap, AT&T set out to design a company-wide assessment program that Field says “really helps you keep track of your career goals and choose the right training.” Launched in early 2018, the effort has two main parts. One, called Personal Learning Experience (PLE), is an online tool that lets employees search for jobs at AT&T based on their current skills, and find out which training they’ll need for their next move.
Someone might, for instance, identify 20 job openings for which he or she is already at least 50% qualified, pick out the ones that look most interesting, and link directly to the training each job will require. Last year, more than 200,000 AT&T employees logged in to PLE more than 7 million times.
One example of how PLE is working is in A.I. and machine learning, where more than 4,600 people so far have had their skills assessed, gone after the appropriate training, and moved from their old jobs into new roles.
The second part, myCareer Profile (mCP), is a tool that’s a bit like an internal LinkedIn page, where employees highlight their job histories, special skills, and which training they’ve pursued so far. Managers search mCP to quickly spot promising candidates for current job openings inside AT&T. The company also uses mCP to get a real-time snapshot of the company’s talent supply, and identify gaps that call for targeted training in specific hard-to-find skills.
You can’t spell training without A.I.
Clearly, none of this, with its billions of constantly shifting data points, would be remotely thinkable without artificial intelligence. But human input is critical, too, in at least three ways. First, there’s basic human nature. To be effective, online assessment tools should be “easy to use, so people will keep using them,” says Jennifer Fitzmaurice, an AT&T vice president in human relations who helped develop the assessments and now oversees them. PLE and mCP, like most of the training they lead to, are available 24/7 on any device “so people can fit this into their busy schedules.”
Then too, the whole effort begins and ends with individual humans, starting with the employees who look for ways to turn their career daydreams into particular goals, and who envision which training will take them furthest in the direction where they want to go.
“People are still responsible for outcomes,” notes Fitzmaurice. “So you need leaders who encourage and inspire people to keep learning and developing.” Encouraging and inspiring are things that algorithms just can’t do.
Rather than using mCP and PLE to make career decisions entirely on their own, and far from letting the online assessments dictate their next steps, the company’s staffers are asked to talk with their managers about their goals, including which training makes the most sense for them.
“We want to make sure employees understand how their current skills can be applied to other jobs,” says Fitzmaurice. “The systems can show you your options. But you always have a choice about what you want to do. It’s still up to you and your manager.”
A third essential role is for humans only: Tweaking the machines’ behavior so they deliver the results the company needs. AT&T’s learning-and-development team, made up of humans, “takes the lead in determining what kinds of assessments and training to offer,” says Fitzmaurice. “But we also gather and use information and opinions from managers in all the business units.”
The process never ends. AT&T experimented with some earlier versions of PLE and mCP in 2015 and 2016 that were themselves extensions of skills-assessment efforts the company had had in place for a decade or more. “We’re still constantly adding and modifying things, partly because the available technology is changing all the time,” says Fitzmaurice. For example, there’s been a recent addition: A self-assessment tool, which more than 200,000 employees have used so far this year.
There’s an added bonus to assessing employees before training them (or even before suggesting what kind of training they might take next): People are creatures of habit and, once they get used to adding new skills they’ve chosen themselves, based on their own goals, they may not want to stop.
Even after 30 years of constant training, Charles Field, for one, doesn’t plan to move to a different job anytime soon—but he’s signing up for more courses to continually broaden his knowledge in machine learning and A.I.
“Machine learning could turn out to be bigger than the Internet,” he says. “I decided years ago I want to be driving the steamroller, rather than standing in front of it.”
More must-read stories from Fortune:
—How this NYU grad landed an entry-level job at Google
—How an entry-level UX designer at Amazon got her foot in the door
—What it’s like to work an entry-level job at Madewell Corporate
—Listen to our new audio briefing, Fortune 500 Daily
Follow Fortune on Flipboard to stay up-to-date on the latest news and analysis.