Science. Ownership. Speed. Openness.
These are the four pillars of Andrew McAfee’s observed structure for successful companies. It is the “geeks,” the leaders at the forefront of cross-industry innovation, who embrace these norms and have the potential to redefine business as we know it. In order to break ground and create the kind of future we dream of, organizational leaders need to banish the fear of failure, embrace mistakes, and accept hard feedback with open arms.
Andrew is a best-selling author, Principal Research Scientist at the MIT Sloan School of Management, and co-founder of MIT’s Initiative on the Digital Economy. His books include More from Less and The Second Machine Age, co-authored with Erik Brynjolfsson. Today on the podcast, we discuss the ideas captured in his most recent book, The Geek Way: The Radical Mindset that Drives Extraordinary Results.
In This Episode
The universal geek (1:35)
The four geek norms (8:29)
Tales of geeks and non-geeks (15:19)
Can big companies go geek? (18:33)
The geek way beyond tech (26:32)
Below is a lightly edited transcript of our conversation.
The universal geek (1:35)
Pethokoukis: Is The Geek Way really the Silicon Valley Way? Is this book saying, “Here's how to turn your company into a tech startup”?
McAfee: You mentioned both Silicon Valley and tech, and this book is not about either of those—it's not about a region and it's not about an industry, it's about a set of practices. And I think a lot of the confusion comes because those practices were incubated and largely formulated in this region called “Silicon Valley” in this industry that we call “tech”. So I understand the confusion, but I'm not writing about the Valley. Plenty of people do that. I'm not writing about the tech industry. Plenty of people do that. The phenomenon that I don't think we are paying enough attention to is this set of practices and philosophies that, I believe, when bundled correctly, amounts to a flat old upgrade to the company, just a better way to do the thing a company is supposed to do. That needed a label, because it's new. “Geek” is the label that I latched onto.
But there's a universal aspect to this, then.
Yeah, I believe there is. I understand this sounds arrogant—I believe it's a flat better way to run a company. I don't care where in the world you are, I don't care what industry you are in, if you're making decisions based on evidence, if you're iterating more and planning less, if you're building a modular organization that really does give people authority and responsibility, and if you build an organization where people are actually comfortable speaking truth to power, I think you're going to do better.
One reason I'm excited about this book is because, you as well, we think about technological progress, we think about economic growth and productivity and part of that is science and coming up with new ideas and a new technology, but all that stuff has to actually be turned into a commercial enterprise and there has to be well-run companies that take that idea and sell it. Maybe the economist’s word might be “diffusion” or something like that, but that's a pretty big part of the story, which I think maybe economists tend not to focus as much on, or policy people, but it's pretty darn important and that's what I think is so exciting about your book is that it addresses that: How to create companies that can do that process—invention-to-product—better. So how can they do it better?
Let me quibble with you just a little bit. There are alternatives to this method of getting goods and services to people, called “the company.” That's what we do in capitalist societies. Jim, like you know all too well, over the course of the 20th century, we ran a couple of experiments trying it a different way: These collectivist, command-and-control, centrally planned economies, those were horrible failures! Let's just establish that right off the bat.
So in most of the parts of the world—I think in all the parts of the world where you and I would actually want to live—I agree with you, we've settled on this method of getting most goods and services to people, most of what they consume, via these entities called companies, and I don't care if you're in a Nordic social democracy, or in the US of A, or in Southeast Asia, companies are the things getting you most of what you consume. I think in the United States, about 85 percent of what you and I consume, by some estimates, comes from companies. So, like them or hate them, they're incredibly important, and if a doohickey comes along that lets them their work X percent better, we should applaud that like crazy because that's an X percent increase in our affluence, our standard of living, the things that we care about, and the reason I got excited and decided to write this book is I think there's an upgrade to the company going on that's at the same level as the stuff that [Alfred] Chandler wrote about a century ago when we invented the large, professionally managed, pretty big company. Those dominated the corporate landscape throughout the 20th century. I think that model is being upgraded by the geeks.
It's funny because, I suppose maybe the geeks 50 years ago, maybe a lot of them worked at IBM. And your sort-of geek norms are not what I think of the old Big Blue from IBM in the 1960s. That has changed. Before we get into the norms, how did they develop? Why do we even have examples of this working in the real corporate world?
The short answer is, I don't know exactly. That's a pretty detailed piece of corporate history and economic history to work on. The longer answer is, what I think happened is, a lot of computer nerds, who had spent a lot of time at universities and were pretty steeped in that style of learning things and building things, went off and started companies and, in lots of cases, they ran into the classic difficulties that occur to companies and the dysfunctions that creep in as companies grow and age and scale. And instead of accepting them, my definition of a geek is somebody who's tenacious about a problem and is willing to embrace unconventional solutions. I think a lot of these geeks—and I'm talking about people like Reed Hastings, who's really articulate about what he did at Netflix and at his previous company, which he says he ran into mediocrity—a lot of these geeks like Hastings sat around and said, “Wait a minute, if I wanted to not repeat these mistakes, what would I do differently?” They noodled that hard problem for a long time, and I think via some conversation among the geeks, but via these fairly independent vectors in a lot of cases, they have settled on these practices, these norms that they believe—and I believe—help them get past the classic dysfunctions of the Industrial Era that you and I know all too well: their bureaucratization, their sclerosis, their cultures of silence. They are just endless stifling meetings and turf wars and factions and things like that. We know those things exist. What I think is interesting is that the geeks are aware of them and I think they've come up with ways to do better.
The four geek norms (8:29)
It's funny that once you've looked at your book, it is impossible to read any other sort of business biography of a company or a CEO and not keep these ideas in your head because I just finished up the Elon Musk biography by Walter Isaacson, and boy, I just kept on thinking of speed and science and the questioning of everything: Why are we doing this? Why are we building this rocket engine like this? Who told us to do that? Somebody in legal told us to do that?
Exactly.
So certainly those two pop to mind: the speed and the constant iteration. But rather than have me describe them, why don't you describe those norms in probably a much better way than I can.
There's a deep part of the Isaacson Musk biography that made my geek eyes light up, and it's when Isaacson describes Musk's Algorithm—I think it's capitalized, too, it's capital “The,” capital “Algorithm,”—which is all about taking stuff out. I think that is profound because we humans have a very strong status quo bias. We're reluctant to take things out. It's one of the best-documented human biases. So we just add stuff, we just layer stuff on, and before you know it, for a couple different flavors of reason, you wind up with this kind of overbuilt, encrusted, process-heavy, bureaucracy-heavy, can't get anything done [corporation]. You feel like you're pushing on a giant piece of Jell-O or something to try to get any work done. And I think part of Musk's brilliance as a builder and an organization designer is to come up with The Algorithm that says, “No, no, a big part of your job is to figure out what doesn't need to be there and make it go away.” I adore that. It's closest to my great geek norm of ownership, which is really the opposite of this processification of the enterprise of the company that we were super fond of starting in the ’90s and going forward.
So now to answer your question, my four great geek norms, which are epitomized by Musk in a lot of ways, but not always, are:
Science. Just make decisions based on evidence and argue a lot about that evidence. Science is an argument with a ground rule. Evidence rules.
Ownership. We were just talking about this. Devolve authority downward, stop all the cross-communication, coordination, collaboration, process, all that. Build a modular organization.
Speed. Do the minimum amount of planning and then start iterating. You learn, you get feedback, you see where you're keeping up to schedule and where you're not by doing stuff and getting feedback, not by sitting around asking everybody if they're on schedule and doing a lot of upfront planning.
Finally, openness, this willingness to speak truth to power. In some ways, a good synonym for it is “psychological safety” and a good antonym for it is “defensiveness.”
If anything, from what I understand about Musk, the last one is where he might run into challenges.
That's what I was going to say. The ownership and the speed and the science struck me and then I'm like… the openness? Well, you have to be willing to take some abuse to be open in that environment.
There are these stories about him firing people on the spot and making these kind of peremptory decisions—all of that is a violation, in my eyes, of the great geek norm of openness. It might be the most common violation that I see classic Silicon Valley techies engage in. They fall victim to overconfidence like the rest of us do, and they're not careful enough about designing their companies to be a check on their own overconfidence. This is something Hastings is very humble and very articulate about in No Rules Rules, the book that he co-wrote with Erin Meyer about Netflix and he highlights all these big calls that he was dead-flat wrong about, and he eventually realized that he had to build Netflix into a place that would tell him he was wrong when he was wrong, and he does all these really nice jobs of highlighting areas where he was wrong and then some relatively low-level person in the organization says, “No, that doesn't make sense. I'm going to go gather evidence and I'm going to challenge the CEO of the company with it.” And to his eternal credit, Hastings goes, “It's pretty compelling evidence. I guess I was wrong about that.” So that, to me, is actually practicing the great geek norm of openness.
So someone reading this book is thinking that this book is wrong. Where would that come from? Would that come from overconfidence? Would it come from arrogance? Would it come from the idea that if I am in the C-suite, that obviously I have it figured out and I can probably do all your jobs better than you can, so why are you challenging me? Why are you challenging the status quo? “Hey, that's how we got here was through a process, so trust the process!”
It's one of the main flavors of pushback that I hear, and it's very often not as naked as you just made it, but it is, “Hey, the reason I'm sitting in this executive education classroom with you is because I'm fairly good at my job. I made some big calls right, and my job is to provide vision to my team and to direct them not to be this kind of lead-from-behind more coach-y kind of leader.” That's one flavor of pushback I get. Another one is a very pervasive tendency, when we come across some challenging information, to come up with reasons why this doesn't apply to us and why we're going to be just fine. It's some combination of the status quo bias and the overconfidence bias which, again, two of the most common human biases. So very often when I'm talking about this, I get the idea that people in the room are going, “Yeah, okay, wow, I really wouldn't want to complete with SpaceX, but this doesn't apply to me or to my industry.” And then finally, look, I'm clearly wrong about some things. I don't know exactly what they are. Maybe the incumbents of the Enterprise Era are going to mount a surprising comeback by falling back on their 20th-century playbook as opposed to adopting the geek way. I will be very surprised if that happens and I'm taking bets like, “Let's go, let's figure out a bet based on that,” but maybe it'll happen. I'm definitely wrong about some things.
Tales of geeks and non-geeks (15:19)
Given what you've said, I would certainly think that it would be easier to apply these norms at a newer company, a younger company, a smaller company, rather than a company with a hundred thousand employees that's been around for 30 years. But it's possible to do the second one, right?
It is possible. Let me violently agree with you, Jim. You and I are of a vintage and we're both Midwesterners. We both remember Arthur Andersen, right? And what an iconic American Midwestern symbol of rectitude and reliability and a healthy culture that kept the business world honest by auditing their books. Remember all that? Remember how it fell apart?
I knew people, and if you got an interview with Arthur Andersen, they're like, “Wow, you are with the Cadillac of accounting consulting firms.”
But beyond that, you were doing a valuable thing for society, right? These people had status in the community because they kind of kept companies honest for a living.
That’s right. That's right. You were true of the truth tellers.
Yeah. It was a big deal and a lot of your listeners, I think, are going to be too young to remember it firsthand, but that company became a dysfunctional, unethical, ongoing, miserable train wreck of an organization in its final years before it finally fell apart. It could not have been more surprising to people of our vintage and where we came from. I tell the story of how that happened a little bit in the book to drive home that cultures can go off track in profound ways and in AA’s late years, if someone had teleported The Geek Way and waved it around, would it have made any difference? I'd like to hope so, but I kind of don't think so.
However, to tell a more optimistic story, I had the chance to interview Satya Nadella about his turnaround at Microsoft, which I think is at a level maybe even above the turnaround that [Steve] Jobs executed when he came back to Apple. The amount of value that Nadella has created at Microsoft in nine years now is staggering, and Microsoft is back. Microsoft has mojo again in the tech industry. But when he took over, Microsoft was still a large profitable company, but it was dead in the water. It wasn't innovating. The geek elite didn't want to go work there. The stock price was flat as a highway for a decade. It was absolutely an afterthought in anything that we care about. And so I use Nadella and I learned from him, and I try to tell the story about how he executed this comeback, and, to my eyes, he did it in a very, very geek way kind of a way.
Can you give me an example?
My point in telling that story is: I do think it's possible for organizations that find themselves in a bad spot—
Established organizations.
Established. Large, established organizations find themselves in a bad spot. Those kinds of leopards can change their spots. I firmly believe that.
Can big companies go geek? (18:33)
What are the first steps to change the corporate culture of a big company?
That's why I'm so blown away by what Nadella and his team were able to do. Let me pick out a couple things that seem particularly geeky to me that he did. One was to say that—it doesn't matter if you develop them or not—you do not own code or data at Microsoft. What he meant by that was, subject to legal requirements and safety and some guardrails, if you want to grab some of the code repository at Microsoft to go try something or some data and go try something, you have the right to do that. That just eliminates huge amounts of gatekeeping and hard and soft bureaucracy and all of that inside the company. And that led to things like Copilot. It's a very, very smart way to start dealing with bureaucracy: just saying, “No, you don't get to gatekeep anymore.”
He also did fairly obvious things like make sure that their really dysfunctional evaluation system was over. He also emphasized this thing that he called “One Microsoft,” which at first sounded like just CEO rah-rah talk. And it is to some extent, but it's also incredibly clever because we humans are so tribal. In addition to the status quo bias and the overconfidence bias, the third easy, easy bias to elicit is “myside” bias. We are tribal. We want our tribe to win. I think part of Nadella's brilliance was to say, “The tribe that you belong to is not Office versus Windows versus Bing versus… the tribe you belong to is Microsoft.”
And he changed compensation, so that it also worked that way. He worked with incentives—he took an Econ 101 class—but he also kept emphasizing that “we are one tribe,” and that makes a difference if the leader at the top keeps saying it and if they behave that way. I think one of the deepest things that he did was act in an open way and demonstrate the norm of openness that he wanted to see all over the place. He got a ton of help with it, but if you talk to him, you immediately realize that he's not this table-pounding, my-way-or-the-highway kind of a guy. He's somebody that wants to get it right, and if you have an idea, you might get a fair erring for that idea. He also embraced agile methods and started to move away from the old ways that Microsoft had to write software, which were out of date, and they were yielding some really unimpressive projects.
So as he and I were talking, I was doing my internal checklist and I kept on saying, “Yep, that's speed. That is science. That is ownership. That is openness,” and just emphasizing, as I listened to him, I just kept hearing these norms come up over and over. But one thing that he clearly knows is that this ain't easy and it ain't fast, and cultural change is a long, slow, grinding process, and you've got to keep saying the same thing over and over. And then I think, especially as a leader, you've got to keep living it because people will immediately sense if what you're doing is not lining up with what you're saying.
One bit that popped out, because obviously I'm in Washington and I see a government that doesn't work very efficiently, and you wrote, “To accelerate learning and progress, plan less and iterate more,” and to iterate means to experiment, it means you're going to fail. And boy, oh boy, failure-averse organizations, you can find that in government, you can find it in corporate America, that acceptance of: try something and if it fails, it's a learning experience. It's not a black mark on your career forever. Now let's go try the next thing.
Exactly. To me, it's the most obvious thing that the geeks do that's starkly different from Industrial Era organizations, “plan less, iterate more.” The great geek norm of speed, and there are a bunch of exemplars of that. The clearest one to me is SpaceX, where they blow up a rocket and that is a win for them, not a loss. And even if it gets written up in the press as, “Oh, Starship blew up, or whatever”—they don't care, right? They'd rather that it didn't blow up or that it stayed together longer, but if they got the learning that they were looking for, then they're like, “Great, we're going to incorporate that, we're going to build another rocket, we're not going to put any people on until we're very, very, very sure, but we're going to blow up a bunch of rockets.” From the start of the company, that has been an okay thing to do.
They also are willing to embrace pretty big pivots. The first plan for Starship was that it was going to be a carbon fiber rocket because carbon fiber is so strong and lightweight, but their method for making it was too slow, too expensive, and had a reject rate that was too high. The thing’s now made out of stainless steel! It's the opposite kind of material! But they said, “Look, the goal is the goal, and the goal is not to stick to the original plan, the goal is to build a great big rocket that can do all kinds of things. The way we get there is by trying—legitimately trying—a bunch of stuff and failing at it with the eyes of the world upon us.”
I want to draw a really sharp distinction between the process and the product, and what I mean by that is a failure-tolerant process can yield an incredibly robust, safe product. We don't need to look any farther for that than the Dragon Capsule that SpaceX makes, which is the only capsule currently made in America that is certified by NASA to take human beings into space. It's how all Americans these days get back and forth to the ISS. NASA doesn't have one. NASA gave a contract to Boeing at the same time it gave one to SpaceX. Boeing still has not had the first crude test of its capsule. This geek way of speed, it's uncomfortable, and you got to be willing to fail publicly and own it, but it works better.
Is the geek way, to some degree, an American phenomenon?
So far.
I was going to say, can the geek way be implemented in other countries? Is there something special about American culture that allows the geek way to work and to be adopted—I said universal earlier, maybe I meant, is it truly universal? Can it be implemented in other places?
Jim, you and I, as proud Americans, like to believe that we're an exceptional country, and I do believe that. I don't believe the geek way only works with a bunch of Americans trying it. I travel lots of different places, and especially the energy that I see among younger people to be part of this transformation of the world that's happening (that you and I are lucky enough to get to observe and try to think about), this transformation of the world in the 21st century because of the technological toolkit that we have, because of the amount of innovation out there, the thirst to be part of that is very, very, very widespread. And I don't think there's anything in the drinking water in Munich or Kyoto or Lima that makes this stuff impossible at all. It is true, we're an individualistic culture, we're kind of mouthy, we celebrate these iconoclastic people, but I don't think any of those are absolutely necessary in order to start following norms of science, ownership, speed and openness. I hope those are universal.
The geek way beyond tech (26:32)
We’ve been talking a lot about tech companies. Are there companies which really don't seem particularly techie (even though obviously all companies use technology) that you could see the geek way working currently?
I haven't gone off and looked outside the tech industry for great exemplars of the geek way, so I have trouble answering this question. But think about Bridgewater, which is really one of the weirdest corporate cultures ever invented, and I haven't read the new biography of Ray Dalio yet, but it appears that all might not be exactly as it appears. But one thing that Bridgewater has been adamant about from the get-go, and Dalio has been passionate about, is this idea of radical transparency, is the idea of openness. Your reputation is not private from anybody else in the company at any point in time. So they've taken this norm of openness and they've really ran with it in some fascinating directions. In most organizations, there's a lot of information that's private, and your reputation is spread by gossip. Literally, that's how it works. Bridgewater said, “Nope. We really believe in openness and everything that's important about your performance as a professional in this company, you're going to get rated on it by your colleagues, and you're going to have these visible to everybody all the time inside the company so that if you start espousing how important it is to be ethical, but your score as an ethical leader is really low, nobody's going to listen to you.” I think that's fascinating, and I think as time goes by, we're going to come across these very, very geeky norms and practices being implemented in all kinds of weird corners of the global economy. I can't wait to learn about it.
I would think that, given how every country would like to be more productive, every country's having a white paper on how to improve their productivity, and this, to me, is maybe something that policymakers don't think about, and I'm not sure if there's a policy aspect to this, but I hope a lot of corporate leaders and aspiring corporate leaders at least read your book.
Well, the one policy implication that might come up is, what happens when the geeks start unignorably beating up the incumbents in your favorite industry. When I look at what's happening in the global auto industry right now, I see some of that going on, and my prediction is that it's going to get worse instead of better. Okay, then what happens?
Save us! Save us from this upstart!
Exactly, but then there could be some really interesting policy choices being made about protecting dinosaur incumbents in the face of geek competitors. I hope we don't retreat into nationalism and protectionism and that kind of stuff. What I hope happens instead is that the world learns how to get geeky relatively quickly and that this upgrade to the company spreads.
The only thing I would add here is I would also urge business journalists to read the book so you understand how companies work and how these new companies that work, companies that look like they are—and not to keep harping on SpaceX, but so many people who I think should know better, will look at SpaceX and think, “Oh, they're failing. Oh, that rocket, as you said earlier, the rocket blew up! Apollo had a couple of problems, they're blowing up a rocket every six weeks!” And they just simply do not understand how this kind of company works. So I don't know. So I guess I would recommend my business journalists to read it, and I imagine you would think the same.
That recommendation makes a ton of sense to me. Jim. I'm all on board with that.
Andrew. This is an outstanding book and a wonderful companion piece to your other work which is very pro-progress, and pro-growth. I absolutely loved it, and thanks so much for coming on the podcast today,
Jim, thanks for being part of the Up Wing Party with me. Let's make it happen.
Absolutely. Thank you.
Thank you, sir.
🌐 My chat (+transcript) with Andrew McAfee on how to upgrade any company like a geek