Welcome to a podcast on the new science of talent, featuring McKinsey leaders and talent experts Bryan Hancock and Bill Schaninger. McKinsey Talks Talent,
Change is a constant. We knew that even in the prelapsarian days before the outbreak of COVID-19. But we didn’t know how fast that change could happen. In the grips of the pandemic, companies and their employees have had little time to devise new ways of working. And the changes will keep coming: artificial intelligence, automation, augmented reality—all the technologies we’ve been talking about in recent years will become fair game. In this episode, recorded just prior to the coronavirus crisis, Bryan and Bill speak with McKinsey Publishing’s Lucia Rahilly about how best to prepare for this new world of work. This is an edited version of their conversation.
For more from Bryan and Bill in the future, subscribe to
McKinsey Talks Talent on Apple Podcasts, Google Podcasts, or the audio player of your choice. Where reskilling works—and doesn’t
Today’s skills, tomorrow’s jobs: How will your team fare in the future of work?
Lucia Rahilly: Let’s start with some basics. Bill, what is reskilling, and how does it differ from on-the-job training and traditional learning programs?
Bill Schaninger: Reskilling has gotten an amazing amount of press, spurred by talk of automation. Will automation and other robotic processes replace people? Historically, you’ve only ever seen that in large-scale manufacturing roles, but now you see automation in roles that previously would have been unassailable—lawyers, medical professionals. Now the question is, “We have a group of employees capable of doing something today. How can we ensure they’re able to do something different tomorrow?”
The answer falls into three categories: “Redeploying,” or moving somebody elsewhere in the company. “Upskilling,” or taking the essence of what employees do and improving it—helping them become more advanced, more gifted at what they do. And “reskilling,” which is old school:
training you in something new. That last one has gotten the bulk of the press coverage because it touches on things like purpose, social responsibility, employer obligation, and employee rights.
Lucia Rahilly: Bryan, do we have research that shows that companies can retrain workers meaningfully at scale?
Bryan Hancock: The research shows companies struggle to train people to do something completely different. What’s critical is to think carefully about the nexus between what a person is doing today and what that person could be good at in the future.
Take insurance claims, for example. Today, a computer can do pretty much everything a claims adjuster can do. From a picture of an auto accident, computer processing can figure out, using artificial intelligence, the damage to the car and how much it would cost to repair—and then spit back out what it thinks the insurer should pay on the claim. That traditional line of business is going away.
So what could insurance-claims agents do in the future? They’d make good sales reps. They’ve seen thousands of accidents, and they can say, “Hey, I know you don’t want that coverage. But let me tell you, I’ve seen when the tree hits the windshield. And your current coverage doesn’t protect that. You really should have it.” So thinking about that nexus—that’s what makes a big reskilling leap successful.
Lucia Rahilly: Are we mostly talking about skills in short supply now, or skills that will be affected five or ten years out?
Bryan Hancock: Both. Some companies are experiencing a shortage now. But if you’re only skilling to today’s shortage, you’re missing what you’ll need in the future. Here’s an example. Futuro Health is a partnership between SEIU-UHW and Kaiser Permanente to train healthcare workers in California. They’re looking at changes in the medical profession and seeing that the more basic parts of care delivery can be automated successfully—but they see a need for higher interpersonal skills. So as they design their training programs, they keep that future gap in mind.
Subscribe to the
McKinsey Talks Talent podcast
Lucia Rahilly: Is there anything surprising about the roles at risk?
Bill Schaninger: The early numbers on automation weren’t all that impressive. Maybe 30 percent of the workforce would have two-thirds of their jobs automated—mostly jobs that were already compromised or were on their way to being compromised. What was surprising was when they broke jobs up into tasks. Suddenly, something like 75 percent of jobs could have up to a third of their tasks automated. That made companies ask, “Are we forced to deconstruct and then reconstruct a job? What does that mean for employees?”
Bryan Hancock: Take a role that’s mostly manual data entry or aggregating information in a spreadsheet, plus a little bit of analysis and presentation. If the basics are automated, just being a star Excel jockey won’t cut it anymore. That person now has to learn how to present, how to think critically.
Often, we think of reskilling as about digital skills. But it’s also about the
human skills that remain once things are automated. And about a lot of the white-collar jobs where dull tasks will be automated; that’s where much of this reskilling and upskilling focus is.
Bill Schaninger: The human element is vital. When you start pulling jobs apart and saying, “Hey, we need you to do these other things,” you’re forced to ask who employees are as people. Do they have any proclivity toward these tasks? What could we reasonably train them on? That’s the first real conundrum: accepting that the incumbents you have—no matter how loyal or even how good they’ve been at their jobs—may no longer be suitable. And that brings all sorts of questions of responsibility to the fore.
McKinsey Talks Talent Podcast
McKinsey Organization’s Bryan Hancock, Bill Schaninger, and other talent experts help you navigate a fast-changing landscape and prepare for the future of work by making talent a competitive advantage.
The case for now versus later
Lucia Rahilly: Most leaders are under huge short-term pressures: Activist investing is up. CEO tenures are down. Geopolitical dynamics are volatile. What do you say to leaders who grasp the importance of reskilling now but want to wait a few years to get moving?
Bryan Hancock: Develop a perspective now on the realities of your workforce. Where will those human beings be in five to ten years, and how can you help them take the next step forward? Some companies, like Walmart, are already investing in their frontline employees—giving them options to learn. I’d encourage leaders to think about those people
now and help them advance.
Develop a perspective now on the realities of your workforce. Where will those human beings be in five to ten years, and how can you help them take the next step forward?
Bill Schaninger: Historically, companies have asked why they’d want to pay for upskilling, given the risk employees might leave. But we have an obligation to make employees employable in the future. Think about a call center where you have 125 percent turnover. Shouldn’t you be even more aggressive about pulling jobs apart and keeping the parts that are actually interesting? When you’ve invested in workers, they’re more inclined to stay. It’s satisfying to them. You can take as a truism that if you don’t do anything different, you’re going to lose people.
Bryan Hancock: If you look at where the biggest gaps are, top of the list is digital skills. But shortly after are skills like English-language customer service. A lot of America’s jobs—at Walmart, at Kroger, at McDonald’s—build skills that are in short supply and will be needed in the future. So this isn’t all about digital. It’s also about core customer-service skills.
Lucia Rahilly: Customer service isn’t a new field. Why would customer-service skills be in short supply?
Bill Schaninger: For a period of time, everything seemed likely to go mobile. Everyone was fascinated with lowering working capital, carrying less inventory, fewer bricks and mortar, fewer people. Now we’ve found a meeting ground: legacy boomers who want to touch before they buy, younger buyers who have a nice interplay of digital and mobile and want experiences, and the group in the middle who do the most on mobile. People are going back to shopping malls, which is remarkable and has made the role of customer service more complicated.
Bryan Hancock: Customer service is also going broader than traditional consumer industries. Healthcare is now much more a customer-service game than it was two years ago. And the roles they’re going to need—to create loyal patients, to coordinate care—rely much more on interpersonal skills than roles they needed 20 years ago.
Bill Schaninger: When hospitals realized they had no choice but to hire for these skills, they started recruiting from hotel chains because they viewed patients as guests—just instead of room service and booze, they’re getting a catheter and an X-ray.
Also, many of us get our flu shot at work, others at the local Target or Walmart. In our lifetime, that role went from a doctor to a med tech who can probably get a certification from a local community college. The actual activity is identical—drawing and dispensing a medicine. But to do it, you no longer need to have “MD” behind your name. As soon as you lower the knowledge or skill requirement for a task, the cost and barrier to entry falls, and more people can do that task. When you pull jobs apart into these component pieces, you find more of those opportunities.
Bryan Hancock: Another area where customer service is getting big is government services. I was recently talking with the chief digital officer of a big US state. The governor’s top three priorities were customer experience, customer experience, customer experience. So when you go to your DMV, you don’t have the experience all of us think of, memorialized in Zootopia. You actually have a representative who is responsive to your needs. The customer-service expectation now goes beyond retail into healthcare, government services, and beyond. That broad reach is part of what’s driving the customer-service skill gap. How to get started
Lucia Rahilly: Do most companies have a proper inventory of the skills currently on their staff?
Bill Schaninger: Not even close. A few years ago, we started looking to invest in how we could pull apart jobs—understand the knowledge, skills, attributes, and experiences necessary to do them successfully—and then try to get a sense of the match between incumbents and what employers need. What became clear was LinkedIn, or services like it—the social platform where people volunteer what they’ve done, where they’ve been, their certifications—that was a better collection of information than most organizations had, because companies collect job titles, not certified skills. That’s been a real disconnect, and many companies are now playing catch-up.
Bryan Hancock: Leaders need to understand the skill supply. Outside-in is the best source of data, but you can supplement that data with good artificial intelligence. Say you’re a McKinsey partner. It’s probably not in your LinkedIn profile that you know how to use Excel and PowerPoint. But AI can infer that we have those skills. So combining outside-in reporting with AI can provide a much better view of what individuals can do. This is where the interoperable learning records come in. If interoperable learning records become more broadly used, you’ll get a much better view.
Lucia Rahilly: Can you give us a quick definition of interoperable employee record?
Bryan Hancock: An interoperable learning record is like a high-school or college transcript, but it gives some of your earlier work experiences the same weight as school. My first job was as a bag boy at Kroger in Peachtree City, Georgia. They called it a courtesy clerk. Cart pusher, bathroom cleaner, “cleanup on aisle five”—all those jobs in one. I learned much of what I use today, in terms of people skills and empathy, from that job. When you see somebody using food stamps for the first time who isn’t really sure what qualifies—you can learn a lot, watching that interaction.
Suppose you categorized the skills you learned as a bag boy in a learning record, alongside your high-school and college transcripts. Together, those would provide a much more holistic profile. And an employer looking for a customer-service or empathy-oriented background could actually reward that frontline work.
Bill Schaninger: As a bag boy, Bryan was paying attention to behavior like, did you struggle with how you paid? Were there opportunities to bag in ways that made things easier for you? That’s part of who Bryan is.
Certified skills are great. But knowing who employees are and then being able to impute what they were likely to get out of an experience is even better. It always comes back to that question: Do you know who your employees are and what you can meaningfully expect to get from them?
Certified skills are great. But knowing who employees are and then being able to impute what they were likely to get out of an experience is even better.
Lucia Rahilly: How do you know that at scale?
Bryan Hancock: If you’re able to capture who employees are—through LinkedIn and artificial intelligence, and in the future, interoperable learning records—you can then assess and score things like conscientiousness. And you can combine those together and impute what the skills are.
The key is saying, “OK, now I understand, at scale, what skills my employees have. But our job descriptions aren’t written with skills in mind.” Organizations like the US Chamber of Commerce Foundation and their JDX [Job Data Exchange™] project are looking to make skill-based job descriptions. But in most cases, if you look at a job description today, you wouldn’t know how to link up employees’ skills with what’s in that job description.
Bill Schaninger: We’re fighting against decades of complexity. We’re trying to get clients to distill the tasks that need to be completed; then you can have an honest conversation about the level of skill actually needed. That constrains or opens a pool up.
Bryan Hancock: Let’s put that into a reskilling context. We did a survey of companies that have undertaken reskilling programs. Those who spent the time to understand the skills required, and their supply, were 1.7 to 1.8 times likelier to have a successful program. Setting up a reskilling program because you’re reading about it and think you’ll need it—that’s the spray-and-pray version of reskilling, versus doing the hard work of translating a job into discrete tasks and linking those tasks up with the skills folks have.
Take telecom companies, for example. If you take employees who worked in the old-school network and move them to an IP-based voiceover internet network, they have to learn a totally different set of skills. Leaders can tell employees, “If you work in the network today and you want the analogous job, then you need a different skill. Eight years from now, this is what our network will look like.” And they can partner at scale to make that transition happen. Or take IT companies that have shifted from mainframes to the cloud. Leaders can pitch to their employees that this similar job requires different skills; it’s a big reskilling, but it’s doable.
It’s harder when the job on the other end is not clear. Your job isn’t transforming—your job is going away. That’s where you have to be thoughtful about which career paths others have undertaken successfully, or which adjacent jobs are likely the most useful. What does a bookkeeper look like, in terms of accounting or other skills? We’re investing in going through literally millions of records to figure that out.
Bill Schaninger: Reskilling starts with asking yourself, What’s critical to our organization being successful? What is that demand in tasks? What is that demand in skills? And what do we have now? Just following that prescription helps leaders get started. You don’t need to have an answer immediately for everything.
We may want every employee to be able to be reskilled, genuinely, because we want to protect them. We don’t want them to no longer fit into our plans. Not everyone is going to be successfully reskilled. But that doesn’t mean everyone doesn’t deserve a choice.