Sutton, Robert

Robert Sutton is Professor of Management Science and Engineering at the Stanford Engineering School, where he is the co-director of the Center for Work, Technology, and Organization and an active researcher in the Stanford Technology Ventures Program. The author of more than 70 articles and chapters in scholarly and applied publications, and coauthor of The Knowing-Doing Gap, he lives in Menlo Park, California. asks ...

Can you tell us about the title, Weird Ideas That Work?

RS: The title of the book is Weird Ideas That Work because the basic premise of the book, which is grounded in many years of rigorous organizational research, is that the logic you need to organize for innovation is dramatically different and nearly the exact opposite of routine work. And the main differences, really, are two. If you want to have an organization or a group of people that keeps breaking from the pack, you need to have variation and ideas that are floating around the group or the company—wide variations—and you need to have people who have what I call vu jadé which is the opposite of déja vu. It's this ability to keep seeing the same old thing as brand new. And for better and for worse, most of what most managers are trained to do—and they should be trained to do, by the way—is to drive out variation and to see the same old things in the same old way.

So most of what happens in organizational life provides people with routine services, making the same product over and over again. And although innovation plays an important role, when it comes to producing a pharmaceutical drug, for instance, you want that drug made exactly the same every single time. Most of what happens in organizational life and where the money is made is on the routine side. We train most people in business schools to drive out variation and see the same old things the same way. So every McDonald's hamburger is just like the one before and somebody visits Disney and they do a ride and it's exactly the same, and so on.

But then, when you expose managers to the kind of techniques that are used a small percentage of the time, and despite all the hype about how everybody should be as innovative as possible all the time—I actually am a disbeliever in being innovative all the time—I actually think for life to work, most of what we need to do needs to be routine most of the time. And it really does. You want buildings to be held up on sound principles. And empirically, most old ideas are good, and most new ideas are bad.

So most managers, most of the time, are managing routine stuff. And when you show them things that are empirically based and are proven to enhance innovation, it looks downright weird. Hence you have ideas in the book such as "Hire 'Slow Learners' (of the Organizational Code)." Why? Because you need people who see the world differently than most people in the company and who bring in varied ideas.

And another one, in terms of rewards—and Tom Peters has been saying this for years, of course—is that you should reward failure. Well actually, I don't think you should reward failure among people who are learning routine tasks like surgery or flying an airplane, because you can identify the difference between success and failure very reliably, and there's a right and a wrong way to do it. And you don't want doctors experimenting on you when they're doing a routine operation, an appendectomy. You really don't.

But then there's the percentage of the time—and the percentage will vary from industry to industry and company to company—where it's necessary to go into this mode of being innovative. And then you have to do things like reward a high failure rate. So I agree with Tom Peters and many of the other writers since who have followed his spirit, that there should be a high failure rate, and it should be rewarded, so long as you want to do something that's innovative.

So that's where the Weird Ideas That Work comes from, is when you show managers who've been successful hiring people who are fast at learning the organization code, punishing or at least not rewarding failure and rewarding success, paying very close attention to what has worked in the past and making little tweaks over it. All that stuff works if you're in the world of maybe a little continuous improvement or doing routine work. But then when you get in the world of actually wanting to do something new and innovative, then they really have to do, very often, the exact opposite of what they've been doing.

And a company I know which is not known for having particularly nice people, by the way, but does this as well as any company I know, is Intel. At Intel, they have two types of work practices. They're into constructive confrontation, meaning they fight over their views like cats and dogs. They teach courses in constructive confrontation, so you learn how to fight and battle out the ideas. That's similar to one of my weird ideas, which is "Find Some Happy People, And Get Them to Fight."

But when they're building their chip plants, Intel's motto is, "Copy exactly," where in every plant, every pipe has to be painted the exact same color in the exact same place. And the justification is, "Physics works," so you can tell if you make a little experiment that things will change. And so, to me, they're saying, constructive confrontation is the logic of innovation, and copy exactly. You're not supposed to argue at all. You're supposed to do exactly what's decided upon, the exact same way in every Intel plant in the world.

So Weird Ideas That Work exists as a book because innovation doesn't really happen that often. I actually think that it's something that doesn't happen that often and often doesn't need to happen.

But it's like the zeitgeist is innovate, innovate, innovate. Half of the books published these days are about managing innovation and making everyone in the workplace a creative person.

RS: First of all, I'm a strong believer in innovation. And if the question is, "What's more important, innovation or routine work?" that's like asking, "What's more important for your car, the transmission or the engine?" They're both necessary. But the fact is by the time we get excited about the new innovations, those things have become routinized products.

Fair enough. We want cars to run, we want planes to stay in the air, we want buildings to continue to stand. You've got to do the routine stuff. Then once in a while, you've got to step out and be innovative. It's two whole different processes. So the big trick here is how do you go along and do the routine stuff, and then switch gears and become very innovative? And how is that possible?

RS: Yes, I think it is possible. We'll start out with the companies that are best at doing routine stuff and know how to step out of it. Then I'll go to other extremes. So you have companies like Intel and Toyota. They're actually both very good about the moments that they need to be, and the places they need to be, innovative. They'll go brainstorm or Intel will have constructive confrontations.

The Toyota production system has, actually, very prescribed things that you're supposed to do—experiments, brainstorming—when you're trying to increase variants and see the problems in a different way. So that would be just making little improvements. But it's clearly the little innovations that keep making the system better and better all the way to one company that we've already talked a little bit about: IDEO Product Development. This is a company that's supposed to do innovative stuff over and over again, and they're mostly living in the world of vu jadé and variance. In fact, the whole trick is to get somebody who sees the product differently, to have as many different design solutions around the same old problem, or to a new problem, as possible.

And there's another guy I know named Joey Reiman, CEO of BrightHouse, which is this wacky design/advertising firm. He's got this concept that his company is really slow and only works on one idea at a time, and they charge clients between half a million and a million dollars for one idea. And the reason that he's able to, if you will, get away with it and have a small but viable business is that he's got variation and he, particularly, sees things differently, by moving as slowly as possible and not rushing through it. And I think that in this world that's fast, slow gives you a different perspective on what you're looking at.

But at IDEO, they're in the business of just being innovative, and they don't have to worry about then turning around and routinizing anything, right? They're an innovation factory. So it seems the routine rules don't apply there.

RS: I spend a lot of time at IDEO. I originally got acquainted with IDEO because I did an 18-month ethnographic study. I was there a day a week for 18 months, and usually was there with one of my doctoral students, who was actually trained as a product designer at David Kelley's [CEO of IDEO (see the interview)]. He was my seeing-eye dog. I've worked with him for about eight years now. I've seen them up close. And even in a place like IDEO, you can identify certain routines that they use to drive the creative process. So that's where things get blurred.

So the work practices they use are very predictable—

Like the deep dive practice, when they go in and—

RS: With the deep dive, the general process is, at the beginning of a project, they look as broadly as possible and they are as wild and crazy as possible about the issue. But they keep getting narrower and narrower and narrower and narrower as the time goes on. Now, it is a routine process for innovation. But still, there are routines.

And plus, they also have to do a whole bunch of things that every company has to do, like they have to pay people. They have to eat lunch. They have to get the building cleaned. So even in a pure innovation factory, which is what I like as a description of IDEO, I would say that's what they are and what Thomas Edison's lab was, as well—there's certain sorts of routines that they use to enable them to innovate.

Right. What's your favorite weird idea in the book?

RS: My favorite weird idea is do something that will probably fail, and then convince everybody around you that success is certain. And this is how the best venture capitalists and the best product development managers work, because it's such a paradox, but it's how Silicon Valley works. It's an incredible model of self-delusion.

The reason you want to do something that will probably fail is that if you're doing something that will succeed, by definition you're imitating something that's proven. When you define innovation, it's going to be an imitation of the past. And the only way to have 100 percent success rate is to perfectly imitate something that's 100 percent proven, and have somebody who's perfectly trained to do it. So, if you want learning, and in this case, innovation, you've got to accept some failure rate. And most venture capitalists, most people who do product development in big organizations, know this.

Mary Murphy-Hoye does a lot of IT research at Intel. She leads a team and she'll criticize people for not having a high enough failure rate. She says, "If you're not going to fail eight out of ten times, we're doing something wrong." I've even heard this from R & D folks at General Motors lately, so it's interesting that that's part of the logic.

In every industry that I can think of, venture capitalists, people who do product development screening, people who pick stocks, I've never seen any academic evidence of screening techniques that really improve the odds when there's massive uncertainty.

At Kleiner Perkins, their ability to predict which of the companies that they fund are going to succeed remains terrible. They're no better than anybody else. Maybe they get a little better deal flow, but they still have an incredible failure rate. But what Kleiner Perkins and other great venture capitalists do, they can create confidence in the outside world that a company is going to succeed, and they can create confidence in the management team that they're going to succeed, to the day that they pull the plug on the company.

And they'll tell you, if they can't express incredible confidence in the management or in the company, then they should pull the plug. So, while knowing that 80 or 90 percent of the deals that they're involved in are going to fail, they're still expressing the greatest possible confidence in their success. Because the most well-proven motivational tool on earth, and the cheapest, is the self-fulfilling prophecy. If someone is convinced that they're wonderful and they'll succeed, their odds of actually succeeding go way up.

I like that because it explains so much of the behavior I've seen in Silicon Valley. I've tried to talk about their failure rate to quite a few venture capitalists. They get upset because they don't want to talk about their failures; they think I'm dissing them.

I've got some emails from two different venture capitalists who gave me estimates of their failure rates. They were estimating about 25 percent.

I had a doctoral student look up the failure rates of their companies—the failure rate defined as companies that were funded by them that never went public, or were never sold for what appeared to be, you know, a decent amount of money to a large company, which is the second most favored option. Empirically, they had a 75 to 80 percent failure rate. But they're getting rich off that, and that's the way the game works. And I think that self-delusion is necessary in that business.

So you think that was self-delusion? Or they were just lying to you?

RS: Oh, I think they delude themselves. But I think they should.

One of your weird ideas is "Hire People You Don't Like." Can you talk about why that works? And following up on that; I know that hiring people you don't like is going to stimulate innovation, but you also say we don't need innovation all the time, so what do you do with this person you don't particularly like who is there all the time?

RS: Those are great questions. First of all, the logic of hiring people you don't like. One of the most well-proven ideas in the behavioral sciences, especially the field of psychology, is the similarity-attraction hypothesis. People who are like us, who agree with us, see the world in the same way, have the same training, the same background, those are the people who we automatically and unconsciously like.

The logic of hiring people who we don't like, who make us uncomfortable, is that they very often will be people who have different ideas than we do. So not liking somebody, having a negative emotional reaction, is very often a sign that you're bringing in people who aren't like yourself into the organization. I don't think you should hire zillions of them, because if everybody hates everybody it'll be, like, too much warfare. But you want to have people who see the world differently.

There are two ways that good managers make this happen. One is they force people into groups who they believe will be disliked. And then they leave them in long enough to drive everybody nuts, and then transfer them out. Or what happens often is there's this curve where they're disliked and they drive everybody crazy, and then what happens is since there's one of them and eight or nine of the others, eventually either the person leaves or starts to be liked by the others.

The other thing that I've seen happen in several companies where they have these brilliant, difficult individuals, who are actually personally disliked, is that they'll find a way to separate this person from the others and just bring him in when necessary. There's one case I know of where one person was really quite widely disliked and pretty nasty to everybody. This guy has the only closed office in an open office environment. And literally, he comes out, especially late in the process of projects, and rips them to pieces. He's probably the smartest person in the company—everybody would agree with that assessment. People go to him for criticism. You don't go to him for compliments. He's very good at finding what's wrong.

But the general point for managers is that if you, in your gut feeling, dislike someone, you have to examine what's causing that. And if it's because they think differently from you and have a different perspective, it might be good to have them around. Although not too many.

We've all worked with people like that. Maybe not as smart as the person you're talking about. The situation I'm thinking about, if it had been managed differently, probably could have worked. But that's your point here ...

RS: The story of this guy's complicated, because he's got bad interpersonal skills. And there's an important difference. What I really wanted to get people to focus on in that case—and this foreshadows something later in the book—is this notion of conflict around ideas. The kind of people whom I think are important to bring in are people who don't agree with the prevailing wisdom and are trained in a valuable but different set of knowledge for the company.

This is the genius of John Chambers at Cisco, as compared to other high tech companies, where they get religious about one set of technologies. He brings in people who have different solutions and different perspectives on building computer networks. And I think that's real smart. And not a popular way to manage. And the natural reaction is going to be to not want those people. So that's where that comes from.

I'm thinking of that famous line from the Apollo 13 movie, "Failure is not an option." That idea seems to permeate America, what with its emphasis on sports and winning. Failure doesn't seem to be very acceptable anywhere.

RS: That's interesting, because I think you put your finger on the irony. And that's one reason you have to keep saying it over and over, that failure is good and necessary, and then it doesn't stick anyway. And this stigma of failure that we have in our culture makes it a lot more difficult to learn from. In management books, for instance, you don't very often read about the way someone screwed up and what they learned from it.

There's a great exception, actually. Don Hastings, who was CEO of Lincoln Electric, he wrote a Harvard Business Review article about all the ways that he screwed up and Lincoln screwed up, in terms of international expansion. I thought that was exceptionally humble and unusual.

There's an interesting dichotomy in the small literature on medical mistakes. There was an anthropological study done on hospital errors. The author distinguishes between hospitals that are good at learning from medical mistakes and ones that are bad at learning from medical mistakes. And the ones that are bad, and therefore make the same mistakes over and over again, have what he calls the "forgive and forget" culture.

That's where you make a medical mistake and you don't talk about it. And the stigma of failure you're talking about, well, in theory, talking about other people's failures is fine. It's our own we tend not to talk about or want people to find out about. We tend to cover it up. But the hospitals that are successful at learning from medical mistakes and make different mistakes and learn from them, they have "forgive and remember" cultures where they're talked about openly. And the people who make the mistakes talk about them openly. And in those hospitals, even when the senior and most experienced surgeons make mistakes, they talk about them.

These institutions have created cultures where it's okay to discuss failures. And that distinction is very important. So the irony you talk about is that all of us management writer types keep talking about how great failure is. But then we throw it into a context where socially it's actually unacceptable to talk about it. We need, somehow, to move towards a "forgive and remember" sort of culture.

You write that creative places are not always enjoyable places to work. Personally, I've always believed in fun. You seem to be saying that fun isn't essential. And then Doug Hall, who at one point said, "It's got to be fun to be creative," is now saying in his most recent book that fun is not necessary for innovation and creativity. He says he's conducted research that came up with that result.

RS: The fun thing is very complicated. Because fun does have value for getting people excited, for keeping them optimistic, and for attracting new employees and attracting clients. So there's a bunch of values to fun in creative work. But it depends what your idea of "fun" is. When I do an analysis of what it really means to do creative work, there's a lot of things that just don't sound like a lot of fun to me. And maybe people have a different idea of fun.

So if you're failing all the time, if you're hiring all these people who you don't like, you're fighting with them—maybe you're spouting ideas, but still you're fighting—maybe some people find that fun. People on the east coast—Boston especially—seem to think that arguing new ideas is great fun.

So there's arguing over ideas. Then, once you learn something, it becomes obsolete; you have to learn something new constantly. If you're doing innovative work, you, by definition, become obsolete much more quickly. And there's a certain comfort in doing sort of the old and the familiar.

And another downside to innovation is that you're constantly having to meet strangers, and so you don't have this history of being understood, of being around familiar people. So to me, there's a lot of reasons why that just isn't that much fun.

And my favorite example of a company is Intel. There are things about Intel I don't like, in that it can be fairly nasty at times. But I have to give them credit. It's a very honest culture. And they took fun out of their core values a little while ago. So it depends, in part, what your idea of fun is. It's interesting that Doug Hall's done research that says it isn't crucial.

You state a couple of different times, talking about managers, that their first law should be "Do no harm."

RS: Yes. Especially in innovation.

What do you mean by that?

RS: Well, first of all, the very root of this assertion is we actually have a lot of evidence—experimental evidence and of course anecdotal—that managers greatly over-estimate the positive impact they have on organizations and products. And also, that we as observers over-estimate how much good or how much harm they're doing. So all the evidence points to the fact that we tend to overestimate the impact—negative or positive—that managers have on an organization.

So we're giving them too much credit, to start. Then we have an entire system where we train managers to do stuff. And then you look at the innovative process, and managers are really good at managing innovation. One person that I'm thinking of is at 3M, which is a company that's always been fabulous at this. What do they do? They give their teams money in a relatively specific direction, and they defend them and they get out of the way.

To the extent that a manager is trained in the traditional ways of doing stuff, they're going to be doing more harm than good when it comes to doing innovative work. So if they're doing the opposite of the weird ideas—

They're trained to meddle.

RS: They're trained to meddle or they're just automatically hiring people who are fast learners, and rewarding success and punishing failure. And the other part related to that is that you can't keep asking somebody who's in creative work how it's going. This guy, Bill Coyne, who was head of R&D at 3M for many years, he had this quote, which I love: "When you put a seed in the ground, you don't dig it up every day to see how it's doing."

That's a great quote.

RS: And for creative work, I mean, one thing that we know is there's this incubation process, and there are long periods of time you actually can't see what's happening. And I think that's an important distinction, say, between McKinsey and IDEO. One reason IDEO typically does not work on-site is that there are these long periods when they're trying to figure out what the heck's going on, and that looks inefficient. Whereas a company like McKinsey, they're always doing something that looks valuable, if you will.

You're a fellow at IDEO. What does that mean?

RS: It's the best worst-paid consulting job I ever had. (Laughs.) I spent about a year and a half doing this ethnographic study, as I mentioned. And working at IDEO is like joining this extended family, and so I got to know everybody well. Then David Kelley said, "We'll make you a fellow, and your job is to find problems that need to be resolved, and then to help me with it. And if I have a question, I'll ask you."

The one job I probably worked the most on with them was the reorganization work. The Palo Alto main headquarters got so big that there would be 120 of us in a room on Monday morning—it was too unwieldy. So it was broken into studios. And the way it was done, it was the best reorganization I've ever seen in my life, in terms of how it was received. I wasn't the only one giving David advice. There were a bunch of us. I don't even know where this idea exactly came from. He talked to Steve Jobs and Jim Manzi. He talked to lots of people with lots of experience.

And what happened was instead of telling people where they had to go, at a Monday morning meeting six people stood up and made a pitch for "Why you should join my studio." And everybody wrote down a first, second, and third choice, but everybody got their first choice.


RS: David is very good at listening and talking to lots of smart people in making a decision; he was talking to a whole group of us for months. And the cool thing about being a fellow at IDEO and working with David Kelley over the years is there's always two things that were kept in mind there: What would be the best optimal structure, practices, whatever, for IDEO; and what would be a management system that would best fit David's personality?

In his case, we're talking about management by staying out of the way. David is the ultimate person who is not interested in details. He's interested in big, strategic stuff, and wants the structure, the organization, and the details to be taken care of, but he doesn't want to deal with them. And so it took a very decentralized structure to enable that.

But he seems to be very smart organizationally. He's great at that. Is he a better organization guy than he is a designer?

RS: That's a great question. Because it raises the issue of a comparison between David and Thomas Edison. The parallels between them are pretty interesting, because like David Kelley—although David's more honest about it—people who worked in Edison's shop would often shake their heads and say that he's known as the great inventor, but he doesn't actually do that much inventing. He does some, but that technically he's not that great.

But Edison would spend most of his time on the organizational issues, and also on selling. Because innovation isn't any good if nobody comes to you and buys your stuff or if that great idea doesn't ever get developed—you have to be able to sell your new ideas when you do innovation.

David Kelley is brilliant organizationally. One of his former secretaries, Gwen Books, said that IDEO was like a toy in his head. It's like a virtual prototype, and he's always playing with it. And he's also a guy who is still very good with stuff. He's always making things. He's very good at product design.

But the salesmanship part of innovation, though it doesn't get talked about much, is crucial. And David is like the all-American boy. He's fabulous at selling IDEO and innovation, and he's very sincere about it. Edison, according to some recent papers, was out there selling the idea of electricity and all the attendant design decisions and other things that were done to get cities and Americans to adopt electric power at a period when there was not a clear technical advantage for electrical power—it was actually quite unsafe initially. And one of the main reasons was Edison was brilliant at selling innovation. So there's a lot of parallels between David Kelley and Thomas Edison.

Thank you.

The Knowing-Doing Gap: How Smart Companies Turn Knowledge into Action (with Jeffrey Pfeffer)
Weird Ideas That Work: 11 1/2 Practices for Promoting, Managing, and Sustaining Innovation