Why Nvidia Wins Where Others Fail & The Truth About Flat Organizations!

0:00Lack of accountability is the root of so many problems in organizations. and we often don't name it that, but when I dissect it, that's what it is. So I am not for super small spans of control because I think that also can cause a lot of problems.

0:27What would happen if we just told the truth? Welcome to TruthWorks, where we dig into the nitty-gritty of leadership and work. and what needs to change. I'm Jessica Neal. And I'm Patti McCord. Our journey together started in HR, but trust us, it's evolved into something wild, honest, and well, a bit rebellious. So throw out the handbook. We're here to redefine rules to work for us, not against us. Let's dive into another episode.

1:03Hi everyone, welcome to another episode of Truthworks. I'm your host Jessica Neal and today I am here. I'm so excited because we're going to take a little different take on the show today. We're going to have a panel discussion on a topic which I'm sure many of you are interested in. flat organizations or should you have hierarchy in your organization? We don't know the right answer, but we're going to dig into it today. And I have one of my favorite people here who has been on the show many times and has co-hosted the show with me. Bob Sutton is here. Hi, Bob. Hi, great to see you and everybody else.

1:41Yes. And another guest who's been on the show before, Beth Steinberg, who we all love. Hi, Beth. How are you doing? Hi, I'm great. Nice to be here again. Yes, it's so good to have you. And Michael Arena is here and he is a first timer. So welcome, Michael. So I should be saying like first time caller. Thrilled to be on the show, but I'm looking forward to it. Yes. Well, Michael, first, before we get started, give us a little intro on you. You know, I know Bob thinks the world of you. And what did you say? How did you describe him a minute ago? He's the most academic person I know who has been a real senior executive multiple times.

2:21He actually can do both, which is kind of rare. Well, you know, and coming from Professor Bob, that's a big compliment, Michael. So tell us about yourself. Why are you so academic? Last time Bob did that, he was a little more straightforward. He said, you're weird. And I said, under most circumstances, I would be terribly offended. But since Bob Sutton said that, like, I'm going to take it as a compliment. It is. It's a compliment. It is. It's a compliment. So that is true. Like I tend to call myself a pracademic. I bridge academia and I'm in academics right now. So I'm a dean of a business school, Crow School of Business here in Los Angeles, right outside Los Angeles.

3:00But, you know, I have been an executive and I've hung my hat mostly in the talent space. you know so I was chief talent officer at General Motors post bankruptcy to help you know really usher in turnaround there which was a blast and I'm sure we will talk a little bit about that but then was it Amazon Web Services as that business was a rocket ship growing never heard of it unlike unlike not in the last couple days either yeah right so that's my background really a talent professional, but mostly I've studied human dynamics. So we'll talk. I'm pretty one-dimensional, even though I span those two spaces.

3:38And most of my research and practice is about collaboration. Yeah. Well, I think to be a great talent professional, I mean, knowing human dynamics is so important. So, you know, but I, you know, I am a talent professional, but not an academic, but I have not gotten the same compliment from Bob about being weird. So, you know, I'm not a bit weird in your own way. You're weird in a different way. I'm weird in a different way. Okay. Well, Beth, remind everybody who you are and where you are today. Yeah. So I'm Beth Steinberg and I've been a long time chief people officer type at many different companies, had senior positions at large companies like Nordstrom and Nike was the first VP of people at Facebook in 2006, done a lot of startups, was recently, most recently the chief people officer at Chime, taking it through their high growth stage.

4:38And now I'm semi-retired. I work part-time for Alphabet, for the Moonshot Factory, helping a company called Tapestry. And I also am an executive coach and sit on a couple boards and advisory boards. So I guess I'm really maybe not semi-retired. So that's what I do. And I actually am a wannabe academic. I try to write and study stuff. So I'm a little bit jealous of Michael being both because I kind of am a wannabe academic. Yeah. Yeah. Well, I'm a wannabe academic too, and I don't think I'll ever be, but that's okay. Same. We don't need to remind people who you are. They know you love me well. I'm an old organizational psychologist.

5:24Anyway. And you're a professor. You are an academic. I'm a professor emeritus. I'm sort of retired the same way that Beth is, I would say. Yes, yes. So let's get started. This topic's amazing. Yeah. So, Bob, you were sort of the idea behind this discussion. You know, tell us why you wanted to talk about this. Well, I wanted to talk about it because it seems like everywhere I go, including one of my colleagues, Rebecca Hines, and I have been doing some interviews about AI in work. In fact, we interviewed both Michael and Beth for it. And everybody keeps talking about how they're in these organizations.

5:58There's many of the people we interviewed where things are getting flatter and flatter. And then I realized that Beth did a brilliant LinkedIn post about it and Michael has actual data. So maybe, Jessica, help me, maybe we can start. And let's start with Beth and then go to Michael. Each of you make a case against overly flat organizations, excessive span of control. So Beth, why don't you go first and then we'll have Michael go. So I wouldn't hear your case. Yeah. So I believe in sort of just enough management, just enough hierarchy. I mean, I think you can go too far in either direction. I think we're seeing a lot of that.

6:43But when I've been in very flat organizations, what I have experienced is employees not really understanding what they're responsible for, not getting the feedback that they need, not feeling like their career is progressing, not really having that sense of trust and community with their leader. And I think that has a lot of long-term consequences for the organization. I think it can also cause a lack of accountability, which I think is, you know, I think about this a lot. I think lack of accountability in my opinion is sort of the root of so many problems in organizations. And we often don't name it that, but when I dissect it, that's what it is.

7:36So I am not for super small spans of control because I think that also can cause a lot of problems. And I'm happy to share some examples of that. But I do think people want to know what they're responsible for. People want to have a relationship with their leader, feel like they work for somebody that they trust and respect. They want to get feedback and they want to have a way to learn and grow. And so, you know, that's why I believe there has to be just enough, you know, structure and hierarchy. And looking at your LinkedIn post, which we'll put in the notes for this episode. Sure. I love your headline, which is managers waking up at 4.30 a.m.

8:21feeling inadequate. So it's hard on the managers who have a zillion. It is. It definitely is. I mean, I think nobody wants to feel like they're disconnected from their, well, in my estimation, maybe some people do. They're, you know, I don't know if they probably shouldn't be a leader, but you, I, one of the, for me, one of the reasons why I stayed being a leader of people for so long is I got a lot of personal satisfaction out of helping people learn and grow and, and building those trusting relationships with my team. and I think it's very difficult for managers to feel like they have no control over that.

9:00And, you know, I do feel like there's a real loss of control when you have 22 people reporting to you. And I don't think that feels good to anybody. Okay, so that's a good one. Okay, Michael, why don't you sort of give your perspective, which is similar to Beth's with some money. I mean, with some data. Sorry, not money. I was like, money? What money? Money. I wish I had some money with it. No, I mean, I agree with everything Beth said. And I think certainly from an employee stance, you know, when you get disengagement, all those things begin to occur. I won't go as far as saying flatter is not better.

9:36I think it just completely depends. And I think it's like everything with management. We tend to be looking for blunt instruments that are universal. And that's almost never the truth. And so we study everything else inside organizations very comprehensively. We look at supply chains. We look at process flows. But we don't study people the same way. So the reality is, and the data says, Bob, that it really just depends. There are some times where you want to go small to go fast. Almost all the work I'm doing right now is on AI adoption. And yet we're not thinking about the burden on managers. And Beth referred to this for sure.

10:18And one of the things we found was as we go in and analytically, and I did some partnership work with Worklytics in particular for this case study, what was found was once you get to like a seven to eight span level on a team, especially a team that needs to move really fast, you start to break the manager. You start adding more people to the manager, and I agree with everything that Pess said about the employees, but just to go to the very problem you're trying to solve, speed, the solution oftentimes actually exacerbates the problem. And what we found was the manager in these cases ended up becoming overloaded and overwhelmed, and they were already overloaded and overwhelmed.

11:01So I think there's just a whole different starting point than saying, should we flatten or shouldn't we flatten? And I think the starting point needs to be, how do we measure speed and velocity of work? And where do we need to move faster? We always talk about augmentation and automation, but we rarely talk about acceleration. And I'm an engineer by undergrad. And there's this old theory called theory of constraints that you look for the bottlenecks and you work on the bottlenecks. And I think a much better approach than just saying, let's flatten and think about spans and levels is to actually just say, where in fact are the constraints and why don't we go solve that?

11:41And in a collaboration manner, which is, again, I confess, the only lens I look at, sometimes what you might be doing is the equivalent of just pushing traffic, opening up another lane on the highway and pushing the bubble farther down the highway. And that's what some of the flattening initiatives do, where leaders are overwhelmed, you give them more stuff, And all you're doing is pushing the decision risk somewhere further down the highway, and it breaks there next. So I think the theory is the wrong starting point. Yeah, yeah, I agree. So, Beth, maybe you can describe this since you're in the middle of Silicon Valley, at least spiritually.

12:23You have so much experience. So why are so many organizations flattening anyways, given all the pain that we're hearing from our friends? I mean, we're hearing stories of pain. I think a couple of reasons. Number one, I do think that, you know, the few years back when things were going gangbusters, too many managers were hired and promoted. And I think that was, and I think, Bobby, you know this, this is one of my big pet peeves is, I think companies did that to reward and retain people. And it had a really negative ramification. I always think about what are the perverse incentives that we're doing inside companies, right?

13:09And I think this was one of them. So I think one of the reasons that it's happening is some of it probably needed to happen. You know, I have examples of even companies that I've been in where you, you know, you take a minute and you look at the structure and you're like, holy cow, we have, you know, 10 managers in one department that have two direct reports each. Well, that really slows when you talk about velocity, Michael, in my opinion, that really slows things down because who's making the decisions and like who's doing the real work. And so, you know, some of it probably needed to be done.

13:47My guess is that most of it's being done as well as just a cost cutting measure, you know, and I think a way to try to make things leaner inside the organization. But I think in many, many cases, we've gone way too far. I agree. And, you know, I think what I see a lot of companies doing, and Michael, it'll be interesting to see if you experience this too. And this is something that we always resisted at Netflix. And I think that you're right. Some of these structures come from a well-intended way of rewarding and promoting employees. But then it adds all of this layers and levels and complexity.

14:37you know you hear these companies and and i'll be talking to a chro and she's like oh you know people people want to know their career path so we need to add levels people want to know like our promotion so we need to add more levels you know and and you know at netflix we kept things very flat so it wasn't that we didn't have any hierarchy we we sort of to your point beth had just enough, it was IC, manager, director, VP, and C-level. We never did the senior director. We never did the senior VP. We never did the senior, senior VP. We resisted that. And we didn't actually have at the IC layer, we didn't have levels either.

15:17We didn't have one, two, three, four, five, whatever. Everybody was just an IC, right? And we called everybody senior, whatever, senior engineer, senior marketer. But we certainly had sometimes from the employees push to do more, but it was like, we're not, because it's going to add complexity into the system that we don't want. Is this going to add complexity into the system that we don't want? And I think that that's where a lot of the complexity starts is that pressure that companies feel to kind of control the chaos of, you know, managers to have an incentive reward program to promote people, reward people.

15:59But to best point, when organizations are thinking about this, they never think, and this is something that Michael said as well, they never think about one, is that something that we want and need, right? And is that how we really want to reward people? They kind of look at what so-and-so, what Facebook did or what this company did, They just sort of slap the same playbook. They don't think uniquely about their own business. And then it's really hard to unwind what you what you have it going. And to Michael's point, it doesn't solve the problem that you're trying to solve. It doesn't accelerate things.

16:35And to best point, in fact, you end up with people who are managing one or two or three people and that slows everything down. So I don't know. I think I've just seen a lot of companies get into a, you know, oh, we got to do this because our employees want it. But I don't think that's the right thing to do. So, Michael, what's your take on this? I mean, again, I think that's why this is such a complex issue is there's an employee issue. There's a manager issue. There's a career issue. There's a business issue. So we almost have to start with, like, what's the core problem we're trying to solve?

17:11Right. And I mean, the way I think about this is form follows function. And like, what's the function? And then the form as opposed to just mandating lower, smaller, you know, shorter spans of control or smaller spans of control if I could get it out of my mouth properly. But the real, I think, you know, so the way I like to think about this from a network standpoint is where do you need speed and where do you need stability? It turns out that flatter and larger teams and driving stable outcomes, you know, where you need to be efficient, where you need to reduce risk, where, you know, where you need to build things that truly work, flatter actually helps a little.

17:51But it turns out that if you're moving fast, you want small, Amazon, we would call them two pizza teams. You want small teams. You want to start to think about the agile approach to work in. And it's almost like you've got to think about what's the function you're trying to solve for and then go after form, as opposed to universally saying that everyone will have spans of eight to ten, and we're going to have no more than three layers. In one place, you may have seven layers. In another place, you may have two. And I just think that we try to solve this with simple-minded solutions, and it's a complex problem, as we've heard here.

18:27There are all kinds of different outcomes and stakeholders. And I think you're mentioning a good point. It's not like a one size fits all. So it's really interesting because one reason I'm so intrigued by both of your perspective on the sort of the great flattening and the problem that's caused is that the contrary cases are these organizations with extreme spans of control and are very flat. So I have a particular obsession. I think that I've dragged Beth into this too, with how does NVIDIA operate? Yes. Jensen Wong has, depending who you listen to, including him, he has between 38 and 60 direct reports.

19:04He doesn't do one-on-ones. He believes everybody should be on the same sheet music. And he does stuff like when he gives feedback to an executive, he does it in front of all 60 or 40 of them. And it's... Sounds fun. And he doesn't have any regular meetings and all this sort of stuff. Yeah. And so I keep asking. I asked our friend Charles O 'Reilly at Stanford how this works, who's a great academic. And I actually was talking to a former colonel about this yesterday, too. And their perspective is interesting. And Jessica, this may have happened at Netflix, too, which is that to the extent you have people who all understand the mission.

19:40So the mission is boss is one of Jensen's phrases. And to the extent so they understand what Charles O 'Reilly was saying, commander's intent, like everybody is on the same page. Yeah. That you can have a flatter organization. Their perspective is that just going in and pulling out five layers, it has always been more command and control. It doesn't work. But if the organization is set up with that logic from the beginning, it may work. So I don't know what Beth or what you think about that, but that's sort of where I'm coming to. And maybe other people have different perspectives. Yeah, I think when you and I first started talking about it, Bob, like that was something that came to my mind.

20:19And Beth certainly mentioned this when she started giving her thoughts on this. If the organization has great communication and clarity and context, I don't think it matters really the structure, right? That is the hardest thing to do is to be clear, to be aligned, to have everybody understand the mission, to your point. If you can pull that off, like that's harder than the structure. And, you know, it doesn't matter whether you have a lot of hierarchy or maybe you're flat. If you have good context and clarity, I think that solves a lot of problems. I agree with that. I mean, I think, you know, Bug did drag me into this a little bit, but I've been very fascinated with NVIDIA because I think when I kind of look across tech companies, it's clear that that company is run in a very special way and that it's run by a very special leader.

21:16And I'm not sure it could be replicated. You know, he has been there, I think, over 30 years. So he really understands, obviously, the company, the product. He is the visionary. He seems, you know, I've never met him. But from what I hear, he is like a good guy. You know, he's a stand-up guy. And so one thing I was thinking about is, you know, there's probably not a lot of tolerance for politics there. information is not power there and you know because he does apparently you know try to share the same information with all you know levels of people and yeah all at once and so you can see how that could work could it be replicated i think it would be very very difficult but i am not skeptical about it i think it works and i think there's very specific reasons why it works.

22:18The weirdest thing is the accountability part and how he does accountability is public performance. At least I just watched an interview with him. And so this I can't believe, Jessica and Michael. So he actually doesn't really like firing people and rarely does. He has this phrase of torturing people into greatness. He thinks everybody can be great, which is a lot different. I do. Good for you. I actually am more of the Charles O 'Reilly school, which is get rid of them fast when they're not working. But maybe that's my problem. So Michael, what's your take on all this madness? I'm wondering if the next podcast is torturing people into greatness.

22:53That was my first thought. I think Beth said this. It all depends on the company. I think a lot of it, you know, we haven't talked at all about culture so far. And I think it depends on like what are the initial conditions? You know, does this work for NVIDIA? I don't know nearly enough about NVIDIA. But yeah, clearly it seems to be working. And the people that I do talk to there are happy. They like it. they enjoy it. It's fast. It's high velocity. It's kind of fits the vibe and the culture. The one thing, Bob, that I thought about was he thinks about scaling one-on-many versus one-on-one. And that's a cultural trait.

23:28If you've got a lot of employees that need high touch, if you've grown up in a culture where people need high touch, they need constant direction or calibrations and check-ins, it would be very hard to operate in that environment. And I think just to be like tactical about it for a second, I think about a pharmaceutical company that we've recently worked with that is very high touch, very relationship oriented. They tried to flatten themselves out. And the reality is that the nice culture of this institution, which if I mentioned the name, everyone would love this company, actually is a liability when you flatten yourself out.

24:06because you don't streamline the decision-making because you still want to keep all your friends and buddies connected in what's coming around the corner. So it didn't change communication a single bit by reducing layers and changing spans. So I think the initial conditions matter immensely, and what might work at NVIDIA or in the average tech company might not work in many, many other companies that have started from a very different place. So I love the example of culture, and there is a structural element to it, too, and I'm curious what other folks think of culture, but at least just in the interview and stuff I've been reading, he says that the 60 or 40 span of control doesn't scale downward.

24:52As you get lower in the organization, the spans of control get smaller. So that's what he's doing, but I'd be very curious to see what it looks like all the way down. So that's another part of the story. Yeah. And it might be, you know, that sort of makes sense that it gets smaller when you go down because maybe that senior most team, that 60 people, he knows really well. And maybe since he doesn't fire anybody, he's worked with them for years, knows how to torture them. To greatness. To greatness. So maybe those relationships are very strong, right? And there's like a trust and a, I don't know, a culture that's bred there that they just have that a lot of organizations probably don't have.

25:42Right? It is, Bob. I'm so glad you added that dimension about deeper in the organization, because, again, I think this is like a speed stability thing as you're thinking about where do you need to go really fast. And we know based on some great work from places like Asana that, you know, as you get larger, you spend more time talking about the work about work as opposed to, you know, being. And the reason Amazon even had two pizza teams or has two pizza teams is because, you know, there's not a coordination tax associated with small teams of people that are synchronized working on a task. They don't need to slow down and add more coordination.

26:19They can just be about the work. And so that's what I was wondering, not knowing a whole lot about how NVIDIA operates is where they're building stuff, where they're inventing stuff, where they're creating new. what do the teams look like there as opposed to where they're scaling, where they're bringing things to market. Those are like two different functions that require different forms, at least in the network. So manufacturing is different than developing the product. So, I mean, it's such an interesting problem. The other thing that I was thinking, the other side of this, and both of you, I've had conversations about this, is one hallmark of companies that avoid this excessive people, excessive hierarchy, is they're actually much slower to hire people and more hesitant to add people?

27:06And so my understanding of Amazon, and I certainly know about Apple, I was just talking to, you know, former Apple executives are like, like they were in the CIA. They were like getting information out of them, even current ones, not let alone current ones, but both Amazon and Apple, what everybody was hiring like crazy or actually hesitate to hesitate to add people, throw people with problems. So I'm curious since you both were, well, three of you were in the talent space. So how do you balance that? Well, we're growing, but how do we not throw too many people or proliferate too many positions as the organization grows, either horizontally or vertically?