The secret to success for many Silicon Valley tech companies isn’t necessarily that they’re ultra-nimble start-ups, or that they’re led by tech-savvy geniuses. In fact, their success often has more to do with a specific type of corporate culture—and it’s a culture that even companies not based on the US West Coast or focused on technology can adopt.
According to Andrew McAfee, a principal research scientist at the MIT Sloan School of Management, business leaders need to think more like geeks, but not the computer-based stereotype you may be familiar with. In his forthcoming book, The Geek Way: The Radical Mindset that Drives Extraordinary Results, McAfee says geeks are nothing more or less than “obsessive mavericks” who are absolutely fixated on finding unconventional solutions to their business’ hard problems. You need them throughout the organization, not just at the top, and you need to entrust them with the power to make real changes.
For this episode of our video series “The New World of Work”, HBR editor in chief Adi Ignatius sat down with McAfee to discuss:
- Evolving a company’s culture not by focusing on organizational structure, but on company norms
- Building organizations that can get things right, even when the person at the top of the org chart is wrong
- The delicate balance of human judgment and evidence, data-driven insights.
“The New World of Work” explores how top-tier executives see the future and how their companies are trying to set themselves up for success. Each week, Ignatius talks to a top leader on LinkedIn Live — previous interviews included Microsoft CEO Satya Nadella and former PepsiCo CEO Indra Nooyi. He also shares an inside look at these conversations —and solicits questions for future discussions — in a newsletter just for HBR subscribers. If you’re a subscriber, you can sign up here.
Adi it’s great to be here.
It’s great to have you. I’ve read the book. It lavishes fulsome praise on geeks in business, not just for their technological innovation, but also for developing an approach to business itself that you’ve come to respect.
Let’s start with a definition. How do you define a geek?
I’m walking away from the computer-based definition of a geek, which was kind of where it started. My definition has two parts. For me a geek is somebody who gets obsessed with a hard problem and is willing to embrace unconventional solutions.
The patron saint of geeks is probably Maria Montessori, who about 100 years ago got obsessed with the problem of how you educate young children best, and came up with the Montessori educational method, which is this radical departure from the industrial scale model of schools that was dominant then, and sadly still dominant now. Think about Maria Montessori when you think about a geek.
How do you define the “geek way” that you’re talking about in the book?
The particular geeks I got excited about that led me to write the book were not educational geeks like Maria Montessori, they were business geeks. They were a group of people concentrated on—but not exclusive to—the West Coast of the US who got obsessed with this problem of, “How do we run a company in an age of really fast technological change and lots of uncertainty? And in particular, how do we avoid some of the classic dysfunctions of the internet era?”
The geek way is what they’ve come up with. It’s the business model, or the culture, or the set of norms that they’ve settled on to try to accomplish that really hard goal.
It’s sort of people who nerd out on, in this case, management, right? Processes of management.
“Obsessive maverick” is my preferred phrase for a geek. And the obsessive mavericks that I got really interested in were the ones who dove in, again, on this problem of, “How do we run a company and keep doing well in our markets and avoid the dysfunctions that seem to plague so many successful companies as they get older and bigger? How do we avoid those dysfunctions? How do we do something that’s higher performing, more sustainable, and a better place to work?”
As I read the book, some of those attributes you talked about, they had to do with speed, they had to do with experimentation. Is this just another term for being digital, or for applying design thinking of the business? Are we talking the same thing here?
No, I don’t think so. Adi, you remember Quibi, right, the Jeffrey Katzenberg-led video startup? It was going to do short-form videos and it was going to change the way we consume entertainment.
Quibi was entirely digital. It was a completely digital enterprise. It was a miserable failure. They raised $1.75 billion. They shut down within 200 days of their launch. It was just a catastrophe. It was a completely digital enterprise.
Netflix is a geek company. They follow all four of my great geek norms of science, ownership, speed and openness. And I think their results speak for themselves.
Geek for me is entirely separate from digital. You can be a non-digital geek, and you can certainly be very digital and not geeky at all.
The Quibi thing is an interesting example. I guess what was missing from your framework was the openness. You had a powerful guy who had been successful, who wasn’t listening.
Not listening to his peers, to his colleagues, to his advisor. He didn’t appear to be a great listener, which is this kind of stereotypical trap that a lot of industrial-era companies fall into. Tech leaders also fall into this trap quite often. You become enamored of your own success and you really stop listening. It’s incredibly common.
One of the things that is powerful, and that I respect about people like Reed Hastings at Netflix is that he was able to build a business that got important decisions right when he himself was wrong about them. In the book I talk about a couple of them.
He was dead-flat wrong about the utility of downloading to the Netflix app. He thought it wouldn’t be very useful at all; it’s incredibly useful.
He was wrong about the importance of kids’ programming for Netflix. And he admits this in the book, No Rules Rules that he wrote with Erin Meyer.
He successfully worked hard on building a company that would correct him, the boss, the high status prestigious person at the top of the organization. It’s part of this great geek norm of openness that I talk about. How do you build a company that will get it right when the person at the top of the org chart is wrong? Man, that’s a hard problem.
It’s a hard problem. And I’m sure some people are going to read your book and think, “Hang on, a lot of these companies we’re talking about are Silicon Valley startups or somewhere in that realm.” The popular perception is these are not cultures that you necessarily want to emulate. They’re often male-dominated bro-cultures. The boss frequently is an overly demanding bully. How do you square all this?
There clearly is some of that going on in the Valley. There are toxic cultures. You mentioned the bro-culture. I would say that Theranos is one of the most toxic cultures I’ve ever heard about, and it was headquartered in Silicon Valley.
I don’t think those are geek companies at all. They might be in the right geographic location, but they’re not following the geek way, which is about these norms and about creating a culture that is fast-moving, free-flowing, argumentative, autonomous, evidence-driven, and pretty egalitarian. That’s the goal of the geek way.
Now you point out some of these companies that I do think follow the geek way that are still too pale, stale, and male. That’s absolutely true. The evidence is pretty clear on this. I hope that gets better over time.
But Adi, you said something that I disagree with, that a lot of these cultures that are not good places to work. That’s true in some cases. Remember when LinkedIn did its top attractors survey, I believe in 2016, they said, look, we’re just going to look at objective criteria. We’re going to look at which company pages get viewed the most often by LinkedIn members. Which companies get the most interest from LinkedIn members, the most applications, and where do people stick around when they take their first job?
The top 11 attractors in the LinkedIn list were all companies headquartered on the West Coast in the industry that we loosely call tech. The geek culture is an extremely attractive culture to work in.
Some of these founders, entrepreneurs who got into whatever they got into because they wanted to make the world a better place, Serge and Larry in their garage trying to systematize our access to all the world’s information, these great ideals: when they are actually running companies, they not only are trying to run a company, but they become killers, wanting to wipe out the competition, to foster monopolistic practices, to just grab as much market share as possible.
Is any of this new? Within the businesses? No, I’m serious. Were the businesses of previous eras cuddly? Did they want their competitors to succeed? Were they trying to rise all boats?
Capitalism is an inherently competitive process. These companies are very, very good performers. If you want to use the adjective “killer” for them, I think that’s by design. You’re not in business not to succeed. You want to grow your market share. You want to grow your profits, you want grow. That is often at the expense of somebody else.
I think it’s for the courts to decide whether they meet the definition of monopolist. It’s a word we toss around a lot. The courts so far have tossed out a lot of the recent lawsuits against some of these giant tech companies. I don’t think they meet the definition of monopolist. In general, for a lot of these companies, the competition is one click away.
And is Tesla a monopolist in the auto industry? You simply can’t make that case. SpaceX has become pretty close to a monopolist in the rockets and satellites industry, but they didn’t start that way, and they’ve become so large and influential because they do the job better.
Let’s talk about companies that are doing it right. This is not universally for digital companies. Some follow the geek way, as you’ve laid it out, some don’t.
How many companies, large, small, digital, otherwise live into these principles, do you think?
I think that’s a really interesting question, and I don’t have a great way to answer it yet. Because the only way that I can think to answer it is to administer a survey to everybody and get them to fill it out. That’s just not going to work.
But you can look at what we know about the cultures at these companies. You can also look at the fantastic Culture 500 research that Don and Charlie Sull did, published in a competitor magazine of yours, Sloan Management Review. They grabbed all the LinkedIn reviews and put them through a machine-learning analysis to see what companies’ own employees said about them.
Three areas I was most interested in were execution, agility, and innovation. And wow, the scores for companies clustered on the West Coast, clustered in Silicon Valley, clustered in industries that we call high-tech, those scores are off the charts. There’s not any real competition for them.
There’s something brewing on the West Coast in industries that we label (for reasons that I don’t like very much) high-tech. There’s something brewing that’s new, that is different than what’s going on elsewhere in the economy, and it’s pretty demonstrably powerful. The label that I hang on that is the geek way.
Was Steve Jobs a geek, and did his Apple follow the geek way or was that a different model?
Jobs has some really classic non-geek characteristics. He believed that he knew best. He had a very, very large ego. He also screamed at his subordinates all the time, which I think is absolutely not what an open leader does.
However, I interviewed Eric Schmidt for the book, and I brought this up to him and he said, “Look, I was on Apple’s board for a while. I knew Steve pretty well.” He said Steve was a tough person in all those ways, but he learned that if you want to stay on top, you have to listen to the people around you. You have to stop thinking that you have all the answers.
We see a really clear example of that with the App Store. Jobs did not want to open up his beautiful, perfect walled garden iPhone to outside developers. He had to be talked into it. He eventually realized that he was wrong about that. There is a little bit of that openness going on.
One thing that Apple is pretty fanatic about, as I understand it, is they make decisions based on evidence. And whether or not that’s a huge AB testing infrastructure is one thing. Apple loves to demo features, and gets everybody in the room to look at this and say, for example, is it better to have blurry portrait photos where you can adjust the blur before you take the picture? The instant they did a demo, they had the answer to that question. They did not sit around and argue from their different points of view. They said, OK, let’s run an experiment. Let’s do a demo here. That’s an extremely geeky approach.
One of the interesting things about Jobs, and certainly something he would’ve said about himself, he did say about himself, was that he didn’t, and that one shouldn’t, slavishly follow the focus groups. People don’t know what they want.
It’s a combination of let’s get input, but there is a lot of personal touch, personal feel. It’s like Moneyball versus the old school. Maybe the answer is a combination of skills.
Absolutely. And Ted Sarandos, who’s the co-CEO of Netflix, says their decision-making is pretty algorithmic. It’s 70-30 algorithmic versus human judgment. But he said, if anything, the human judgment is on top, if that makes any sense.
I think that makes a ton of sense. This norm of science that the geeks are pretty passionate about, it’s not just dry analysis and do whatever the data says. That’s an inaccurate caricature.
Science is about settling arguments with evidence. It’s a ground rule for how you’re going to come to a decision. You’re going to do a test, you’re going to do an experiment. You’re going to sit around in a group and argue about it. And if you can’t agree, evidence is going to settle this issue.
Science is this inherently, deeply social, deeply argumentative process with a ground rule. Evidence is the queen here. That’s what we’re going to follow. And a lot of the geek companies that I respect do that as naturally as breathing.
One of the examples in your book of geek culture triumphing is Microsoft, and specifically when Satya Nadella came in and rekindled that early success and more. Talk a bit about what he got right, particularly in the framework that you’ve created.
Can you think of a more impressive corporate turnaround story in living memory?
No. Microsoft went from the tech company we liked least to maybe the one we like best.
And if you were an investor, you really didn’t like it for about a decade. The stock price was flat as a corpse’s EKG. And then Nadella took over, and it’s become one of the most valuable companies in the world. It’s an astonishing comeback story.
I got to interview Nadella for the book. He’d made brilliant strategic moves, a bunch of them. You have Microsoft embracing open source, wow.
But I was interested in the cultural changes that he made. And he did a couple things that I think are straight out of the geek playbook. Some of them are obvious. He embraced agile development methods as widely and quickly as possible inside Microsoft. He did a couple things to reduce the sclerotic bureaucracy that was in place at Microsoft, which was just hamstringing their ability to do anything important out there in the world.
One of the brilliant things he did was say, “Look, you cannot own a digital resource inside Microsoft. You cannot own the data. You cannot own the code.”
And what he meant by that was you can’t be the gatekeeper. You can’t say yes or no to other groups who might want to use it. With that one simple move, he said to the company, “Look, if the AI group wants to go grab all of the GitHub code to train up a model to help provide assistance to programmers, you don’t have to ask permission. Just go do that, subject to all the right constraints and safeties on it.” Man, that is an astonishingly good bureaucracy-reduction mechanism.
Maybe the deepest thing that Nadella did was that he pulled off this amazing feat of helping Microsoft become a less defensive organization. What I mean by that was, he said in his interview with me, “We had a culture where it was not OK to be wrong, to show any weakness, to not hit your numbers, to not be the smartest person in the room.” He said, “We just had that culture and it had to change.” He did a number of really brilliant things to move from a culture of defensiveness to a culture of openness.
When I used to hear this word, “vulnerability”, in connection with leadership or business, I thought it was just a buzz phrase, hand wavy, nonsense business. The company is not a therapy group. It’s not there so you can sit around feeling vulnerable all the time. I was just wrong.
Now, the company is not your therapy group. However, a successful company needs to be a place where it’s okay to be wrong, to fail, to not have the answer, to show that you’re uncertain. Nadella helped get Microsoft down that path, and it was an absolutely fundamental thing to do.
I also interviewed Yamini Rangan, who’s the CEO of HubSpot here in Cambridge, who took over a culture and has strengthened it through a really difficult time through the pandemic.
She discussed one of the things she learned and that she was good at was saying in this unbelievably uncertain time of the pandemic where tech companies were shrinking. It was all weird. She said to a lot of her constituencies, “Look, I don’t know. I don’t know what the future holds here. I’m going to be honest with you.” She also shared her board performance review with her direct reports, not just the good parts, but the stuff that she needs to work on too. These are all just great moves to start to show the rest of the organization it is okay not to be perfect, not to put on the brave front, not to be winning all of the time.
Jack Welch’s autobiography was called Winning, and it epitomized this industrial era view of what you have to do all day every day. I love the geek view, which is, “Hey, man, we’re going to launch some rockets and they are going to blow up. Now, we’re not going to kill anybody, but we’re absolutely going to launch some rockets that are going to blow up on the launchpad.”
Bezos said a few years back in the shareholder’s letter, “We are incubating multi-billion dollar failures inside Amazon right now. That’s appropriate for a company of our scale.” And you look at Alexa and I think maybe he was right about that.
But the point is that this obsession with winning and being on top and being right and being dominant, that has to go away.
I want to get to audience questions. One came from Shabana in Pakistan. What kinds of organizational design, organizational structures, do you need to foster this sort of geek culture?
I don’t think org structure is the key because the companies that I surveyed have very, very different org charts. They also have very different formal practices.
Netflix is fairly famous for having the no-vacation-days policy. Amazon has extremely strict vacation policies for different levels of employee.
I think it’s not a matter of the org chart or the org structure that you have. It’s not so much a matter of how formal a lot of your policies are. It’s a matter of your norms. I love that word and I use it in the book all the time.
A norm is a behavior that the people around you expect. Maybe it’s written down in the employee manual. Very often it’s not. It’s community policing. It’s what the people around you expect. If you go out of line and violate a norm in a community, you will know that fairly quickly and you will either come back into line or you’re just not going to stick around very long.
If you can work on these norms of science, argue about evidence, of ownership, push authority and decision-making down to an uncomfortable degree, speed, iterate, don’t plan, don’t analyze, build stuff, get feedback, learn from reality and then openness. Don’t be defensive. Be willing to pivot. Be willing to admit that you’re wrong. They show a little vulnerability. Those are the norms that are critical for the geek way.
So Bob from our audience is asking, “Is the stack of books on your right your reading list for the week?”
These are from all over the place. But there was a stack of books that I kept referring to when I was writing The Geek Way, and they were not business books. I am sorry to admit this as a business book writer.
They were books from this relatively new field called cultural evolution, which gets at this fundamental question, “Why are we the only species on the planet that builds spaceships?” Nothing else is even close. We’re really the only ones out there. The octopuses are not going to do it. The ants, the bees, the chimpanzees are not going to do it. Why are we humans the only spaceship-building species on the planet? This field of cultural evolution to me has come up with the best answer to that question, which is we are the only species that cooperates intensely with large numbers of people that we’re not related to. We are the species that learns the quickest, that improves its toolkit, its technologies, its cultures most rapidly over time.
You can take that and put that to work in a company. A company is a large group of mostly unrelated people. And the goal of a company is to improve its culture, its artifacts, its technologies, its practices over time. The goal of a company is to practice very rapid cultural evolution. Now that we know a bit about how cultural evolution happens, we can put those insights to work.
There’s this massive unexplored opportunity to take the insights from this field and put them to work inside the company. I think it’s so massive because I haven’t heard anybody talk using cultural evolution’s terms inside even very geeky companies. This is very, very new stuff. And I think The Geek Way is the first applied business book of cultural evolution.
I think I just figured something out. You’re a total geek.
We’ve been working together for years. It took you this long to realize this?
But the thing that I got obsessed with was that in my career, I’ve spent a lot of time in the (I hate these labels) old economy and the new economy, lots of industrial-era incumbents, and then a lot of companies clustered in the tech space, clustered on the West Coast.
For my whole career, I have been trying to pattern match and figure out what the deep differences are between the two. Is it the foosball tables? Is it the hoodies? Is it bringing your dog to work? No, obviously not.
Where the action is in the economy, where the excitement is in the economy, what are they doing differently? And The Geek Way contains my answers to that kind of career long, geeky, head scratching question.
Let’s get practical for our viewers who aren’t in classically geeky cultures. How can they apply the approaches you’re championing, at their own companies?
Here are a few things you can start doing tomorrow that are extremely geeky practices that will get some traction. When you’re in charge of your group or you’re a team member or you’re leading a team, if you say things like, “Oh, that’s a good point. I hadn’t thought of that. Let’s go that way,” or, “Oh, I was wrong about that. That’s really good to know,” you have just demonstrated openness and vulnerability instead of classic defensiveness.
That’s hard to do. Human beings are inherently defensive people, we don’t like being shown to be wrong in public. It’s deeply uncomfortable. But if you model that behavior, it will start to spread. And I’ve seen geek leaders do that over and over.
You can start to push decisions and responsibility down to a place that makes you uncomfortable, and to get out of saying, “Just run that by me first,” or, “Make sure I’m included on that,” and try to get rid of the hard and soft bureaucracy that gets in the way.
I was talking to Sebastian Thrun, who’s an alpha geek in all kinds of things, and he said, “One of the things I try to get my teams to do is stop so much communicating.” He had this great image. He said, “A team works on something, and then they run it up the flagpole.” He said the really natural tendency is for the people at every layer of the org chart to add something to that because they want to be part of the solution, they want to be part of this winning idea. By the time that idea gets back down to the originating team, it’s unrecognizable. Get out of that business.
Again, this is uncomfortable stuff to do, but it’s a big step toward the geek way. And then finally, you can start to say, “How are we going to make this decision? Can we agree on what evidence we’re going to look at to make this decision?”, as opposed to saying, “Well, I’m the boss,” or, “I got it right last time, so you should listen to me this time,” all these other ways we have to make decisions, credentials, hierarchy, charisma, excellent PowerPoint, all these things: get out of that business and start asking for the consensus on, “How are we going to settle this argument? What evidence are we going to look at?” These are all super geeky practices you can start doing tomorrow.
Here’s another audience question. This is Lori from North Carolina, and it’s another very practical question: how can organizations get leaders to unlearn the idea that failure is bad?
This is a super hard question, and I fall back on something that I learned from Clay Christensen: one of the things Clay taught me is that managers are voracious consumers of theory. He said, “Don’t be afraid to tell them why this needs to change, what the principle, what the framework is.” And he said, “If you want to get people to change, there’s a kind of a tripod. There are three things you have to do over and over again more until you’re absolutely exhausted.”
First of all, tell them a vivid story. We humans respond to narratives. Second, give them some evidence, give them some numbers to show that it works. And third, give them a theory. Explain why it works. You’ve just got to go through all three of those legs.
“Why can’t we make decisions this way? I’m a pretty smart person. I reached this place on the org chart because of my excellent judgment. What do you mean we should be relying a lot less on my judgment?” Again, three part answer to that question, keep iterating on it. It is not an easy process.
I think a lot of companies who get interested in what Silicon Valley does differently or might want to embrace the geek way are going to find it hard going because a lot of it is unnatural and uncomfortable. Challenging your boss in a meeting in most circumstances is uncomfortable, until it becomes a norm, until it becomes what people around you expect.
I can’t let you go without talking a little bit about generative AI. You have an article in the next issue of HBR on how businesses can most effectively capitalize on the technology. What is your thinking, your quick overview, on the utility of GenAI, and its likely impact on, let’s say, the white-collar workforce?
The more I think about GenAI, the more I think we’re underestimating its impact. It is so general, it is so powerful, it is improving so quickly, it’s leading to so many complementary innovations. I have never seen the computer science geeks, the economic geeks, and the business geeks all get so excited about anything in my career, which is actually a quarter-century long by this point.
Generative AI is going to change the business world. It’s probably going to do it quicker than a lot of us are thinking, and waiting around on the sidelines and not being part of that transformation is a deeply lousy strategy: get a plan together and then just go do stuff.
All right, Andy, we’ve run a little bit long, so we’re going to cut you off here. I could talk all day with you about all these topics, and our audience is clearly into it and engaged. Thank you very much for being on the show.
Adi, always a pleasure. Thank you for having me.