Middle management has been under a decades-long assault—and in the COVID-19 pandemic, the crisis-driven need for speed has turned “flatter, faster, leaner” into a mantra. But what’s at stake longer term?
In this episode of McKinsey Talks Talent, McKinsey leaders and talent experts Bryan Hancock and Bill Schaninger speak with McKinsey Global Publishing’s Lucia Rahilly about the vital role many middle managers play in their organizations—and about what we might stand to lose in a flatter future. An edited version of their conversation follows.
For more from Bryan and Bill, subscribe to the McKinsey Talks Talent podcast on Apple Podcasts, Google Podcasts, or the audio player of your choice.
What’s changing—and what’s more of the same
Lucia Rahilly: Bill, Bryan, here we are, nearly a year into COVID-19, and we’re all aching for this pandemic to be over. Today we’ll talk about rapid changes in the way companies have organized during this crisis—specifically, changes to the traditional management hierarchy as more leaders pivot toward a flatter, more distributed, more agile organizational structure.
Bryan, let’s start with you. Why are leaders making this organizational shift?
Bryan Hancock: During the early days of COVID-19, leaders made decisions incredibly quickly and with a small number of people—even on big strategic shifts like opening new digital channels. Senior leaders began to say, “Hey, we just made decisions in two weeks that used to take two years. Can’t we think about speeding up the organization all the time?” And then they shifted to, “Maybe we can just live with the set of people making the emergency decisions. A lot of folks weren’t in that meeting. Maybe we don’t need them. Maybe we can actually have a flatter, faster, leaner organization.” But while enticing, the notion of eliminating whole swaths of middle management and emerging flatter and faster—we might need to pump the brakes on that a bit and consider the nuance of what middle managers do most productively in regular times.
Bill Schaninger: What COVID-19 presented resembles what you might feel after a natural disaster—the need to act quickly to protect lives, livelihoods, the institution. I think what we’re confusing is a response to COVID-19—a natural tendency to massively concentrate power—with thinking there’s not a benefit to having middle management distributed throughout the organization.
Lucia Rahilly: Have you seen the ranks of middle managers decrease?
Bill Schaninger: The ranks of middle management have been under a 25-, 30-year assault. If you were to roll back the clock to Hammer and Champy, Reengineering the Corporation: A Manifesto for Business Revolution [Harper Business Books, January 1993], that work triggered a nonstop look at what we’re getting from the middle and whether, at each point, leaders could have a wider and larger span. We’ve seen endless rounds of cost cutting, an ongoing erosion in confidence that what middle management does is value creating. And certainly in 1990, there was a lot of low-hanging fruit. But after 20-some years, many organizations have hit the point where the connective tissue has begun to fray.
What further clouds this now is the rise of artificial intelligence, machine learning, statistical techniques—tools and approaches that when deployed well, are great decision aids. But at the extremes, people suggest these tools obviate the need for humans to make decisions. And that’s why you still see a push toward asking, “Hey, do we really need those middle managers?”
We’ve gone a bit sideways there, wanting to wire out the role of the human. We’ve allowed the environment to cloud the purpose of a well-performing individual in a well-structured middle-management role. We have to tease that out, not oversimplify it, not overextend the experience we just had from COVID-19, which is not going to apply all the time. It’s an artifact of what we’ve been enduring.
The difference a good manager makes
Lucia Rahilly: That’s an interesting point on AI [artificial intelligence] and decision making. In a distributed structure, who upskills frontline workers to make independent decisions and function productively on cross-functional teams? Has training happened to enable this on teams not operating this way prepandemic?
Bryan Hancock: You’re hitting on a key element here: the role of the manager in training and people development. Take, say, a leader of a technical competency, whose role is to define what good looks like, shape learning journeys, shape career development, aggregate year-end feedback—not to be a bottleneck but to adopt the mindset, “What I’m here to do is to coach, support, lead.” When managers spend the majority of their time coaching and leading, we see real returns. And during COVID-19, people are starting to realize that extra manager time really makes a difference.
Subscribe to the McKinsey Talks Talent podcast
Bill Schaninger: Think of the role of the direct manager in performance management. Ultimately, it’s not about the system. It’s not about the form. It’s about whether your direct manager helps you understand the value of your work and how it fits into a broader strategy. It’s about coaching. It’s about real-time feedback delivered by a human. I think we’ve undervalued those elements, tried to systematize them away with data, with systems, with tools. Meanwhile, workers are asking for better leaders, better apprenticeships, better coaching. When we try to disintermediate the role of a good midlevel leader, I think we do this at our peril.
We’ve also been down this road before—it’s not actually that novel. In the early ’90s, some US automotive manufacturers had successful autonomous or semiautonomous teams. They often ran plants, or at least teams, collectively. Teams understood the economics of what they were doing. And midlevel managers served as the glue between different agents, different constituencies, which really made it work. But when leaders rotated out and didn’t start with the same premise, the same commitment to the model, things often fell apart. That didn’t mean that suddenly autonomous or semiautonomous teams were a bad idea. It meant suddenly you had a different midlevel manager who was either not sufficiently committed, not sufficiently trained, or didn’t have sufficient social capital to pull it off.
Lucia Rahilly: How much investment has there traditionally been in learning and developing social capital and other leadership capabilities at the middle-management level?
Bryan Hancock: It varies. The vast majority of corporate training covers onboarding, compliance, and the rollout of new products or systems, with relatively little on leadership development. Sometimes organizations recognize the importance of the leader and offer training programs for key transition periods. Then a gap might emerge: once you’ve undergone that transition, are you getting sufficient ongoing professional development to help you become a better coach and leader? That ongoing learning-and-development cycle is where we see less funding, less support, and also where we see some innovation coming.
Bill Schaninger: During COVID-19, we made everything virtual, and that’s an interesting opportunity to remind people of the difference between training and learning. Training often focuses on the job you’re doing in the moment. You can break it up into bite-sized chunks. Learning is often experiential, often has a social component, and is often a bigger part of the fabric of your organization. At McKinsey, historically, we’ve had signature moments—what happens when you’re nine months in, when you’re 18 months to two years in. Big events where, yes, training and development is important, but the spirit of a cohort, the spirit of camaraderie, the spirit of recommitment to the firm is also important.
What COVID has challenged is the importance of bringing people together. I think we’ll learn over time that this is not a budget exercise around training—you can certainly deliver training cheaper. It’s about investing in your culture and in setting norms about how leaders behave and what’s expected of them. We probably have to keep that kind of learning separate, otherwise we’ll lose out to the budget crunch.
McKinsey Talks Talent Podcast
Lucia Rahilly: And how do frontline workers develop if middle-management opportunities dwindle?
Bryan Hancock: A counterintuitive example of a middle manager, somebody who sits exactly midway in the hierarchy from the leader of the overall organization and the frontline, is a school principal. A principal usually has assistant principals, division heads, and teachers below in the hierarchy, and a regional superintendent and a superintendent above—almost the definition of a manager in the middle of the hierarchy. And very few people would argue for a wide-scale elimination of school principals. If anything, principals play an incredibly important role in engaging with parents and the broader community, as well as in leading teachers and ensuring they develop and deliver the best for students. Plus, people recognize that good teachers want to work for good principals.
There’s something we can learn from applying that example to other industries. Just because you sit somewhere in the hierarchy doesn’t mean you’re not valued. And in fact, the better you are as a midlevel manager or leader, the more impact you have not just on developing good frontline folks but also attracting them in the first place.
Investing in the leaders of the future
Lucia Rahilly: What about the next step? How do you assess the leadership pipeline without middle management serving as a test bed?
Bill Schaninger: In many of the cases we’re discussing, the common denominator is the broad distributed organization—physical plants or locations that are various and spread out. An individual might be the senior-most person for a location but in the middle of the organizational hierarchy overall. That’s a great example of an opportunity to serve as the face of the organization, translate everything to the local context, and provide day-in, day-out frontline-level coaching and guidance. It’s a great test bed for finding out who your future leaders are. Back in the day, the famous label was the “popcorn stand”: you gave someone an early experience of P&L [profit and loss] responsibility and people leadership to see whether they had the competence and the potential to go much higher in the organization.
I think middle management is part of the fabric of your leadership pipeline. These roles should be coveted and nurtured and curated, not eliminated. If you want to eliminate something, eliminate tasks—tasks that are administrative or bureaucratic and don’t add value. But keep the role and curate it to help develop your next generation of leaders.
Lucia Rahilly: Many leaders have made new commitments to diversity in hiring and leadership, especially in the aftermath of the killing of George Floyd and the global protests against racial injustice. Does culling middle-management positions pose a risk to development trajectories for racial minorities or women, who tend to be underrepresented at the leadership level?
Bryan Hancock: There’s a falloff in diversity, going up the hierarchy, for African Americans and Latinos and also for women. And yes, if you aren’t thoughtful about how you define middle-management roles, you’ll have difficulty developing a diverse pipeline. That’s not just about the number of these roles; it’s about what people do in them. Are they administrative, in which case they may actually become career dead ends? Or are they the kinds of roles we’ve been discussing: people leadership roles, roles that involve applying judgment or influencing part of the P&L? If you craft middle-management roles the right way, you create opportunities for the people in them to thrive. You also create capacity for them to coach and mentor diverse people below them, so that the next level can rise as well.
Middle management is part of the fabric of your leadership pipeline. These roles should be coveted and nurtured and curated, not eliminated. If you want to eliminate something, eliminate tasks—tasks that are administrative or bureaucratic and don’t add value. But keep the role and curate it to help develop your next generation of leaders.
What’s right for your organization?
Lucia Rahilly: Bill, give us your best quick and dirty on spans and layers—the way organizations determine the right “span of control” for each manager and how many reports that manager should have. Suppose you’re a leader looking across your organization and you’ve seen a leaning out at the middle. What should you be thinking about?
Bill Schaninger: First, assume you’ve done it wrong previously, and I mean that in the best way possible. Normally, when people have gone through spans and layers, they solve to a number: “I need to get this to at least seven.” And they start changing the nature of the role by creating an economy of scope—taking the role they want to keep, or often the person they want to keep, and bolting on extra tasks or responsibilities. They end up with a Frankenstein. The structure no longer makes sense or lines up with how the organization makes money.
So instead of starting with spans and layers, start by grounding on the basics of a corporate identity. What’s your purpose? How are you going to make money, not just today but in the future? How do you want to run the place? What kind of place are you? Once you anchor on the answers to those questions, you can ask how you should organize to get the work done, and what the right distribution of leadership of responsibility might be.
If you’re an academy company, for example, and growing leaders is your business, you’ll likely push responsibility deeper into the organization, which means you’ll have general manager roles, often economies-of-scope roles with broader spans. If you’re an execution company, you might prioritize frontline leaders engaging in coaching and apprenticeship, in which case you might artificially constrain their spans to create time for those priorities. The point is that the spans you assign ought to come out of the role—the type of role, the archetype of the role, how it fits with your culture, what you’re asking people to do. Not solving for a number.
Lucia Rahilly: You mentioned purpose there—obviously a hot topic during the pandemic. Many leaders are talking about and revisiting purpose and the need to serve a broader group of stakeholders than shareholders alone. What’s a leader’s stake in helping middle managers find a way forward?
Bryan Hancock: Over the next decade, we expect to see a lot of automation in areas where early managers and midtier managers learn the basics of their trade—how to review a particular line in a financial statement, for example. As AI takes over those tasks, that training ground won’t necessarily exist. You’ll have a gap, where someone fresh out of high school or community college will need to make the leap to a more supervisory role, or one that requires more judgment, more insight.
So the first level of middle management may actually require thinking and working alongside machines, while also being able to exercise judgment that normally comes with experience. How do you get that? How do you build to that? And as a leader, how do you reorient what employees do so that you’re positioning the next generation for success? The need to come up to speed faster ups the necessity for training, mentorship, and development in that middle layer.
If you craft middle-management roles the right way, you create opportunities for the people in them to thrive. You also create capacity for them to coach and mentor diverse people below them, so that the next level can rise as well.
Navigating the next phase
Lucia Rahilly: Suppose you’re a leader, and you’re in the middle of an agile transformation that may have been accelerated due to COVID-19. How can you help anxious middle managers make that shift?
Bryan Hancock: I’d start with a conversation about expectations for their role in the new world. Put very simply, “Here’s the role you’re in. Here are the responsibilities of the role. Here’s who you interact with. Here’s what authority you have and don’t have. Here’s what we expect you to achieve.” And then talk very specifically about jobs they’re expected to do over the next three months—leading this product or ensuring this team has this particular set of capabilities ready to go. Middle managers should walk out of the conversation with clarity on what you want them to do, so that they shift from something nebulous to something very tactical that increases their chance of succeeding.
Bill Schaninger: For sure, help them get clear on what really matters about their role. And also help them look at all the things they’re currently doing, because a lot will be historical, or personal preference, or normative—“just how we do things around here.” A lot of that should stop, in order to create capacity—capacity to do a job well, to spend time on leadership and apprenticeship and coaching, to spend time on their own personal training and development. Those aren’t incremental activities.
Lucia Rahilly: And how does creating capacity fit with objectives leaders have looked to achieve during the pandemic—reducing complexity, driving faster decision making, but also curtailing costs?
Bryan Hancock: Getting very clear on the role and the jobs to be done helps with speed and cost. Organizations can move faster when there’s clarity about what a person’s job is and who that person interacts with, when there’s no ambiguity as to who is responsible for which decision. And having that focus also confers a personal reward. Because at the end of the day, people want to create that new product, deliver that amazing new curbside pickup service innovated during COVID-19, not run up and down a hierarchy getting signoffs. Role clarity helps with speed, helps in aligning with purpose, and it helps with cost when you identify the jobs to be done, in part because it also helps illuminate the jobs that should no longer be done.
Lucia Rahilly: Does it help with pandemic fatigue also? Making time for managers to learn and develop would seem particularly important as the crisis continues and organizations begin to tire.
Bryan Hancock: Absolutely. Middle managers need to be thinking about how they personally energize, as well as how they energize the folks around them.
Bill Schaninger: One of the biggest ways to help people is not to treat their job as a ratchet that only ratchets up. We add more and more, but very rarely do we help people set things down. Some jobs have become overwhelming, almost undoable—and then you add on middle managers’ belief that their days are always numbered due to annual cost cutting.
Bryan Hancock: There’s something even about the term “middle management” that denigrates the value people bring in leading the assignments of teams, leading a school, leading a plant, leading the learning and development capability within HR. Those are all examples of what some people would call middle management. But they’re vital—much more important than the stereotype that, unfortunately, the middle-management term might conjure.