The science and psychology of upskilling: why learning by doing is so important

“Like jumping from a bridge.” That was how one employee at French telecoms operator Orange described a skills programme that has helped old-school engineers to master cloud-based technologies, while managers retrain in roles such as cyber security.

Thankfully, Orange does not just wait for people to sink or swim; it gives them a safety net. “When we offer a worker the chance to change jobs, it’s not six months in the classroom,” says Vincent Lecerf, executive vice-president of human resources and group transformation at Orange.

Instead, the individual joins a new team — on their existing salary and the promise of continued employment — in an approach that, from day one, interleaves on-the-job learning with formal instruction and coaching. “What makes the difference is not just the quality of the training but of the mentor — the feeling that as a learner you’re supported,” Lecerf adds.

Orange’s approach frames upskilling as a shared responsibility — something that organisations often struggle to achieve. Compliance-style training, on the one hand, ticks a box, enabling workforces to sprint through the basics — but at the expense of relatable content and genuine impact.

At the other extreme, on-demand learning promises choice and the freedom to go at one’s own pace, but puts the onus on workers to take the first step. Company “learning libraries”, which gather resources in one place, have a mixed record. These promise to put life-long learning at employees’ fingertips but have struggled to attract users and show returns. Although the need to upskill is there, most “people are so busy . . . that only if [learning] affects the job they’re doing right now will they get involved”, says Daniel Godfrey, head of learning impact at Hemsley Fraser, a UK-based training provider.

Fortunately, humans are not dependent on the latest kit to upskill. With the right culture, they can learn from each other and often very effectively, as a body of evidence shows. In his 2013 book Social, Matthew Lieberman, a psychologist at the University of California, urged educators to encourage their students to lift their grades by teaching others what they learn. 

One theory on why teaching boosts learning is that explaining something pushes the brain’s retrieval mechanism to try harder, alerting us to knowledge gaps and bolstering what we have grasped, making it easier to recall. A further possibility, which Lieberman favours, is that simply knowing our efforts help others stimulates the part of the brain devoted to forging social ties, which spurs us to work harder. 

What helps students can also benefit workplaces. Lee Holmes, chief executive of Infinox, a fintech and trading platform, asks his dealers and relationship managers to sit down and swap skills. “As a service-based business, if the relationship manager is away, our dealers need to know how to speak to our clients in layman’s terms and be a person, not a robotic response . . . and it’s good if the relationship manager understands the dealing side,” he says.  

When you hear something that forces you to think about an issue from a new angle, it can lift your whole afternoon

To build people’s skills, Spark Foundry UK, a Publicis Groupe media agency, holds lunchtime debates on topics that affect advertising, such as the rise of influencers. The debaters are randomly assigned a perspective to research and the whole agency is invited to hear them out over pizza. The framework adds peer pressure on participants to put in the work and is an incentive for those listening to pay attention. “When you hear something that forces you to think about an issue from a new angle, it can lift your whole afternoon,” says Kate Anthony, Spark managing director.  

While tackling the unfamiliar can feel daunting, the good news is that learning is pretty much what the entirety of human evolution has wired our brains to do. That said, to get the most from learning “we need to protect our brain and know how it functions”, says Chiara Succi, Turin-based associate professor of organisational behaviour at ESCP Business School. For a start, “sleeping is definitely underestimated”, she adds. While the body sleeps, the brain is busy processing information, forging connections and laying down memories.

While training used to consume a whole day, “we now know that breaking things into smaller chunks is more beneficial”, says Natalie Mackenzie, a brain injury and cognition rehabilitation specialist. That allows learning to be put into practice and reflected on. In experiments, mixing other tasks with the practising of what has been learnt — for example, combining learning a new IT system with serving customers — appears to play a pivotal role in consolidating knowledge. This is where the brain shifts new information from working memory, where it must be laboriously recalled, into areas housing memories that kick in readily.

Similarly, putting learning to the test stimulates the brain to spot patterns, derive principles, retain what is useful and discard what is not, grasp where our efforts fall short, and apply what we learn more creatively by making wider associations. 

Research pioneered by psychologists Robert Bjork and Elizabeth Ligon Bjork suggests that for knowledge to stick, learning must be effortful and allow for getting things wrong. 

But what of occupations where a simple mishap could put profits or even safety at risk? This is where advanced technologies come in, enabling “workers in high-risk roles to learn from their mistakes without actually doing harm”, says Susan Wade, a learning and development expert at professional services firm PA Consulting.

By solving a problem you’re going through the thought processes . . . you’re building that muscle memory

At Immersive, a UK-based cyber resilience company, clients are thrust into mock cyber attacks to conjure up the adrenalin rush of confronting the real thing. For senior leaders, a typical format might be: “You’ve been hit by ransomware attackers — it’s in the press, what should you do next?” says James Hadley, Immersive’s founder. 

AI is used to simulate footage of news reports of the attack, or journalists calling executives asking for comment, while panicked customers flood social media demanding answers. This intensifies the pressure as participants decide what to do, and the crisis simulation dynamically unfolds.

If that sounds a lot to handle, it is designed to be. “By solving a problem you’re going through the thought processes,” Hadley says. “You’re building that muscle memory because when you do get hacked, you want to be ready to respond.”