According to Deloitte’s John Hagel, the best collaborative teams are diverse and built from the bottom up.

MIT Sloan Management Review: John, we talked to you about 18 months ago. Is there anything new in your research that you’re particularly excited about?

John Hagel: We just completed a major research initiative around business practice redesign, but the whole focus was a bit contrarian. For several decades, the whole thrust in the business world has been around business process reengineering. Our view is that processes are increasingly becoming prisons and the key to accelerating learning and performance improvement is actually at the level of small frontline work groups that are confronting new situations. We did a series of case studies around work groups that are accelerating their performance improvement and identified nine practices that seem to be contributing to that acceleration.

Do you see companies moving toward a work group structure as opposed to the traditional command-and-control hierarchies?

We call the work groups the invisible part of the organization. Companies are relentless in tracking performance at the level of individuals and of departments. Companies know work groups are out there, but when we asked them if they have any work groups that are accelerating performance improvement, we got kind of a blank stare: “Well, I’m not sure. We have work groups, but we don’t really know what performance improvement they’ve achieved and whether it’s accelerating or not because we don’t look at that.” It’s interesting that more and more of the performance at companies hinges on these frontline work groups, and yet they’re invisible.

Is there a difference between work groups and cross-functional teams? And if so, what is the distinction?

The way we defined a work group is a group of people who spend the vast bulk of their time together. They’re not just the team that meets once a week for a couple of hours to coordinate. And the work they do together is so intertwined that you can’t pull it apart and say, “You do this step first, you do this step second.” It’s not a process. It’s collaborating around addressing a particular problem or an opportunity that requires anywhere from three to 15 people.

Are they cross-functional? It depends. How broadly do you want to define cross-functional? Some of the work groups we looked at worked in maintenance at Southwest Airlines Co. They brought in people from other functions into the work group, but it was largely maintenance people. On the other hand, we looked at an emergency ward in a hospital. At one level it’s a function — there’s an emergency ward staff — but they come from very different disciplines or backgrounds in terms of their medical specialties. So, it varies. Typically the high-performing work groups we looked at each had people from different kinds of backgrounds or experience sets.

Do you think diversity is a requirement?

Absolutely. Actually, maximizing the potential for friction was one of the practices we found in successful work groups. That goes completely counter to most large companies’ culture, which views friction as bad. We want to be efficient. We want to do things as quickly as possible. Friction within the team slows us down.

A key part of success is what bringing together people with diverse perspectives, experiences, and backgrounds engenders. There are a whole range of ways of looking at diversity, but the key was to bring together people who weren’t just going to automatically agree on what needed to be done. They were going to come with different ideas and perspectives, and challenge one another.

Do you see a commonality in who’s starting these work groups and companies, and how they’re forming?

It was largely bottom up. It was not the CEO saying, “We need work groups with these practices.” It was frontline people coming together saying, “We need help in addressing challenges. Can we come together as a group to do this?” And improvising on the fly. They didn’t have a process manual that said how to form a work group and how the work group operates. They were inventing it as they went.

Are other skills helpful for working in a work group other than this friction concept?

We had nine practices in total. We don’t describe them as skills per se, because skills tend to be defined as specific activities that are context specific, like being able to program in Perl or to operate a specific kind of machine. These practices are independent of the specific context, though they need to be tailored to the context. For example, one practice is framing a powerful question. These work groups were driven by a question they had framed that nobody had an answer to but that would change the game if they could come up with an answer. And they were driven to constantly try to find ways to answer that question.

You described this almost as grassroots. If these work groups have such benefits, and if they’re largely driven by people getting together when they see a need or a question or a common goal, what do organizations need to do to encourage this?

Where do I start? Everything needs to change, frankly. Part of the reason we focused on work groups is more and more head counts at large companies are spent on what we describe as exceptions. They occur when companies try to automate and standardize processes, but increasingly there are events and things occurring around them that the process didn’t anticipate. So, you improvise as you go.

I think it means getting out of this mindset that everything is a process. And it’s a bit of an embarrassment, because large companies spend hundreds of millions of dollars, sometimes billions, on automating processes. And to actually acknowledge that there are more and more things happening that a process can’t handle — whoa, not sure I want to acknowledge that. We call it the shadow economy at large companies, and it’s not going away. It’s a fundamental trend. Part of the “big shift” we describe is that the world is becoming more uncertain, more unpredictable, and you need to organize in a way that allows you to address that. Processes do not allow you to address that. They put you in a prison.

It means challenging organizational silos, the way we organize in a hierarchy and keep everybody in their areas and reward them only if they stay in their silos. These work groups are kind of off on the edge, if you will, and not really recognized by the company as a focal point for performance.

You talked in your report about the difference between high-performing teams and work groups. Can you talk a little bit about that?

Again, this is a bit of our contrarianism. There’s an infinite amount that’s been written about high-performing teams. Regarding the question of what’s a team versus a work group, we have a higher bar in terms of work groups being groups that work together on a continuing basis, full-time. They’re not just there for a meeting. But the other piece that’s probably more important is this notion that, in the big shift, where there is mounting performance pressure, “high performing” is a static concept. It’s looking around and asking, “Are they performing better than others? OK. Tomorrow they’re going to have to perform even better.” And what are the specific practices that are required to not just improve performance but accelerate performance improvement over time?

Part of our view is that there is a big shift in that we’re in an exponential world where we don’t just need linear performance improvement but accelerating performance improvement. Our sense is work groups are going to be where that learning and performance improvement occur. It’s not going to be in a training room or when you listen to someone lecture. It’s going to be adopting a set of practices that help you get better faster. And it’s not enough to view yourself as a high-performing work group. One of the things we found in these work groups is they were never satisfied. Even though they were high-performing by any definition, they wanted to get better and faster, and their focus was on how to do that.

But you talk about work groups not being permanent. How do you continue to build on that performance if you have to disband that group at some point?

Well, the work groups we looked at are pretty long-lasting. Some of the examples, like the emergency ward or the fire department, create pools of people and pull them in to form a work group for a specific challenge or issue. Somebody came in with a particular illness, or a fire is happening down the block. Who can we mobilize? Those work groups were constantly changing in terms of the specific participants, but the pool was relatively stable. A set of people were constantly coming together and learning together through the work group activity.

Who should be in charge of measuring the performance of the work groups?

That’s a great question. One issue is that companies tend to be focused on financial metrics, and our view is that the performance that matters the most is actually the operating performance of things you can measure on a front line and on a very regular basis, not the financial performance. In my view, it’s somebody in the operating part of the organization. But the challenge is that they’re siloed. So, you’ve got a silo over here — they’re measuring performance in their silo — and a silo over there — they’re looking for performance in their silo. Who is responsible for measuring these work groups and how they’re performing is an open question.

Have you seen examples where a company is doing it better than others, specifically in a way that’s helping the work group continue to advance and maybe even send signals up through the chain to leadership? It seems like this is fertile ground for sensing and learning about what’s happening and how the company needs to adjust and adapt.

Our hope with the research we did is to create some visibility for this, because even the process of finding these work groups was a nightmare. We would visit the top executives in the company and ask, “What work groups are accelerating performance?” Again, we’d get blank stares. We couldn’t find any examples of companies that really embrace these work groups and are trying to spread the practices to other work groups in the organization. This is very much a bottom-up, undercover thing, because these work groups are actually violating a lot of the processes that the silos have defined, so they don’t want to talk about it. They just want to get the job done and get to a higher level of impact.

Do you acknowledge them, to bring them out of that state of invisibility and to try to measure that?

Our hope is that with companies under increasing performance pressure, we can show them parts of their organization that are accelerating performance improvement and ask, “Wouldn’t it be powerful to scale that and have those practices embraced in more work groups within your organization?” Hopefully, the way we get the attention of senior executives is by focusing on the performance improvement that’s been achieved and say that’s something that’s more and more important to the company, and it’s time to focus on this invisible part of the company.

I realize it hasn’t been out for too long, but who is showing the most interest in the report? In terms of roles within the organization, is it the executive level or the operating level?

I’d say it’s more of the operating level at this point. To be fair, we haven’t really aggressively gone out to try to engage senior executives on this yet, but that’s part of our game plan. But I think you actually just saw the first version of the report. A second version was just released that goes into the nine practices in much more detail, with a separate section for each practice. Then the third report will be a collection of the case studies where we go deep into these individual work groups and show how these practices were manifested in the work groups’ actions.

You talked about having better business practices rather than processes. What do you mean by that? How do companies do practices instead of processes? Is it connected to this work group concept we’ve been discussing?

I’m generalizing, and I’m sure there are exceptions, but I would say most companies pay zero attention to practices. It’s all about the process manual. Specify the activity, designate who’s going to do what in what sequence, and standardize it to make sure it’s done in exactly the same way everywhere and as efficiently as possible. The way performance is typically defined in the process world is around efficiency. Faster, cheaper. Our view is that if you are truly going to accelerate performance improvement, you have to stop focusing on efficiency. If it’s just efficiency, that’s a diminishing-returns game. The more cost-effective and faster you are, the harder it’s going to be to get to that next level of efficiency. But if you focus on effectiveness, on impact, on value delivered to whatever the arena is — the sky is the limit. That requires a mindset shift, getting out of that efficiency mindset.

Is that because technology is changing at an exponential rate, which means the opportunity to impact, to create different practices that have a greater impact, is increasing exponentially, too?

Yes. We’ve made a chart that is a simple comparison of the improvement in price performance with some digital technologies and labor productivity. To be fair, labor productivity has improved over the last few decades. It’s more than doubled. But over the same time period, the improvement in price performance in technology has gone up exponentially by many orders of magnitude. And so, to us, it highlights this gap between the potential of the technology and the actual application, the way people are using the technology to deliver value. That’s the big white space and the opportunity areas to figure out, again, what practices can help people accelerate performance improvement.

So your hope is that organizations will see the benefits of these work groups and scale them?

That’s an interesting challenge because it’s not a process manual that you can write down step one, step two, step three. And one of the challenges with practices is the actual activity that’s done within a practice is going to differ dramatically. We looked at the whole notion of friction in a fire department, where if you’re fighting a fire, you can’t, in the moment, debate the best way to fight the fire. On the other hand, in other kinds of activities, there is time to actually debate before you act.

How the practice manifests is going to differ depending on the context, but it’s important to at least highlight what the practices are and why they seem to have such an impact on performance. We can point to some examples and say, “By the way, this isn’t just theory. Some work groups are actually doing this, and here’s what we’ve learned from them.” And making that more visible and spreading the word to the other work groups in the company.

This really ties into the broader concept of learning. Can you talk a little bit about the importance of learning in this day and age and how learning in work groups might be different from what we’ve seen before?

Oh, don’t get me started. This all goes back to our view of the big shift, which is, on the one hand, creating an environment of increasing performance pressure. Everything’s going to be harder and harder over time. There’s going to be more demand for performance, and if you don’t rapidly improve your performance, you’ll be marginalized. On the other hand, the paradox is expanding opportunity. It creates more and more potential for performance improvement, and it’s in that context that we’ve made learning a centerpiece of our agenda for executives. But the challenge is that when most executives talk about learning, it’s about training programs: what program at what point in the career of the person? And the really edgy people around learning are talking about how to deliver training programs to the workplace in little video snippets so that workers can listen to the lecture while they’re confronting a situation.

Our view is that the most powerful learning in this exponentially changing world is not sharing existing knowledge; it’s creating new knowledge on the job, in the moment, in the work environment. And we need to create environments that promote that, and practices to manifest that. That’s where the power is.

Most work groups are not accelerating their performance improvement. We’re looking at only a very specific subset of work groups that have adopted at least some of the nine practices and are accelerating performance improvement. But one of the reasons we focus on work groups is our belief is that no matter how smart you are as an individual, and how much you can learn as an individual, you’re going to learn a lot faster as part of a work group. This happens particularly, again, in a diverse work group where you’re being confronted with different perspectives and being constantly challenged, and you get better faster.

If a group does create new knowledge, is that then spread to other work groups or scaled across the organization, or does it mainly stay within the work group? Is it creating instead of sharing knowledge, or is it in addition to sharing knowledge?

No. It’s certainly not about creating knowledge while ignoring the need to share it. But in my experience talking to executives — and again, I’m going to generalize — nobody is talking about creating new knowledge unless they’re talking about the research and development function in their organization. If you happen to have some researchers in a lab somewhere, they’re creating new knowledge, right?

Their innovation lab out in Silicon Valley does it.

Yes, exactly. But the rest of the organization just needs to do what they’re told and learn from the teacher and the training sessions.

How do they go about creating new knowledge? How do they work differently to make this happen?

The focus is at the work group level and the practices. All nine of these practices come together to shape and accelerate that learning. Part of it is creating friction but also creating environments where people can challenge one another with respect. You’re not doing it to put one another down. You’re doing it because you have to have a common objective and an aspiration to get even better.

One of the practices has to do with reflection, regularly stepping back and analyzing what you did. Very few people do that. In a world where we’re under more pressure to do things faster, who has time to sit back and reflect, much less as a group? It’s the combination of these practices that really drives the opportunity for accelerating learning and performance improvement. It’s learning through action. Another key piece of our perspective is when we talk about learning, most people think about sitting in a room, listening to a teacher, or reading a book. No. It’s learning through action, learning by doing something, and seeing what the results are and refining your actions over time based on the results you achieve.

Would it be fair to characterize the learning by action as experimentation — trying new ways of doing things, and then seeing what works and iterating from there?

I guess. I just get allergic to experimentation, because that tends to be viewed as something that’s done on the side. God forbid we should experiment when we’re maintaining an airplane or fighting a fire. But on the other hand, it is the notion of constantly trying new things with a goal, being driven to higher and higher levels of performance. It isn’t experimentation just to see what happens. It’s trying to get to a higher level of impact than has ever been achieved before.

Do these work groups have formal or informal leaders? And does it take different leadership skills to lead these work groups, to create new knowledge, than has traditionally been in the workplace?

Yes. Most institutions today operate around a model of scalable efficiency, and the mark of a strong leader in a scalable efficiency environment is someone who has all the answers. No matter the question, you can count on that leader to have the answer. And God forbid if they say they don’t have an answer, because that’s a sign of weakness. Fire that person. Get somebody who does have the answers.

The mark of a strong leader in what we describe as a scalable learning world, which is what we believe we need to move into, is the one who has the most powerful and inspiring questions, who will freely acknowledge that they don’t have the answers and will actually ask for help. That creates an environment that inspires people to learn faster, because they have a question they don’t have an answer to, but wow, if we could answer that, we’d be really successful. And it invites everyone to ask more questions.

If you have all the answers, you’re not asking tough-enough or good-enough questions.

You don’t even know what the questions are. A part of the reason we believe trust is eroding in our institutions is the whole facade that leaders present. We know we’re in a world where nobody has all the answers. So, if you’re saying you have all the answers, guess what? I don’t trust you. Either you’re clueless or you’re not being honest with me.

Is that hard for leaders to come to grips with? Is that mindset shift a particular challenge in organizations adopting this sort of learning or work group mindset?

It’s a huge obstacle. It’s a huge cultural barrier. Again, saying you don’t have an answer is a sign of weakness. You’re going to get fired.

And that means we ask questions we know the answer to, because then we’re sure we’re not going to get fired.

Yes. Then we’ll look really good because we have the answers. Here’s the question and here’s the answer. Now go do the work.

You’re talking about work groups and pushing toward learning, and we talked a little bit about experimentation, too. What is the difference between experimenting to try to innovate and work groups that focus on performance improvement?

That moves us to a whole other discussion. I’ve become so tired of the word innovation and the way companies approach innovation. It’s something that’s done off on the side — we’ve got an innovation lab in Silicon Valley, but everybody else just does their job. Our view is that everyone in the organization needs to innovate, and not just occasionally, but every day they’re on the job. Another thing we’re contrarian about is the belief that it’s possible to design ambidextrous organizations, where part of the organization focuses on innovation and the rest focuses on execution. No. Everybody needs to constantly challenge themselves about whatever they’re executing on, asking themselves, How could we do this better? And not just in terms of efficiency, being faster or cheaper, but to get better and better results, more values delivered to whoever their constituencies are, and that’s missing.

One of the answers I get from executives when I talk about scalable efficiency and scalable learning is, “Great! That’s really exciting. We’ll do both.” And our view is that you have to choose. These are in fundamental conflict with each other, and one illustration of that is the question, What’s the message you get in a scalable efficiency world? The message is failure is not an option. You deliver as forecasted without exception. What does learning require, especially learning in the form of creating new knowledge? It requires failure. If you’re not failing, you’re not learning fast enough. There’s a fundamental, cultural conflict between those two goals or orientations. With ambidexterity, do we just silo off the innovation? Well, there’s going to be some failure and some experimentation that’s going to be done over there. Everybody else does their job and executes without failure.

Organizations are having a hard time serving their customers and delivering their core products and simultaneously trying to innovate. What advice would you give to companies that are trying to do both? If a car company is looking into how to innovate and think about the future of mobility at the same time, how does it continue to support customers who want its current model?

Actually, customers don’t want your current model. They want a better model, and part of the reason companies are failing is that they’re not constantly bringing out better and better products and enhancing the customer experience in every dimension, everything from when you walk in to buy a car to when you have a service problem to just operating the car on a daily basis, all of that. The customers are becoming more and more demanding, and if you just have a group on the side that’s talking about the future of mobility and autonomous cars, and all these wonderful things, but you just keep producing the car you have, that’s not going to work.

So, your answer really is that the whole organization should be doing this, even those who are delivering what the company would consider their core business?

Exactly. Everyone. I go down to the janitor in the building facilities. How can we make the building facilities a better experience for the work groups that are in them? How can we make the food service a better food service? It’s constantly trying to help people, whoever your audience is. Some parts of the company deal directly with the customer, like sales and customer service. But every function — finance, human resources, IT — should be striving to get better and delivering more value to whomever they’re serving within the organization or outside the company.

We’re in a world of mounting pressure, and if you don’t get better, you’re going to get marginalized. Our research around return on assets for all public companies in the United States from 1965 until today has collapsed. It’s eroded by 75%, and it’s been a sustained erosion. No sign of it leveling off. No sign of it turning around. (Just in case you want an indication of the imperative to learn and the broken nature of our existing practices.) It’s not that we’re not trying harder in our companies. We are squeezing harder and harder to get more efficiency. That’s not the answer. The answer is to create more value everywhere, every day. That’s the missing piece.