How to screw up less when it matters most: Olivier Sibony

A perfect storm for risk aversion

“Basically organizations are a machine to produce risk aversion under the control of the hierarchy and the benign control of the social group around the people who should be taking risks.”

What is it like to be a business?

“Organizations don’t think. People think. And organizations operate in various ways. So we need to find a way to get from the individual level to the organizational level. We’ve got a level of analysis problem here. And my read of this is that organizations fail when they fail to overcome the cognitive biases of individuals.”

The problematic “best practices” approach

“I can’t tell you how many times the client, when I was a consultant, has asked me, “How does that company solve this problem?” And that company was usually Apple or Google or Toyota or some company that was admired for whatever reason but which had nothing to do with the company in which the client who was asking me the question was sitting.”

Not getting fired on the way to improving your company

“A lot of people who read my book or who attend my seminars say, “I’m going to go tell my boss that the way we make decisions on his team is not good.” I really do not recommend that you do that. You will not win friends and you will not influence people. This is not going to work and it might actually end your career faster than it should end. So start with your decisions. The focus here should be on how do I improve the quality of my decisions and my team’s decisions and the decisions that are under my control. It’s not about “you have biases”. It’s about, “I have biases and I want the decision-making process to be a safeguard, an antidote to my own biases”.”

The balance of powers in a healthy choice structure

“A good organization has a decision-making process that doesn’t just let people come up with whatever idea seems right, which is what we all do as human beings. It actually has some checks and balances in the process and organizations fail when they fail to have the checks and balances that they should ideally put in place.”

Brooke: Hello everyone. And welcome to the podcast of The Decision Lab, a socially conscious applied research firm that uses behavioral science to improve outcomes for all of society. My name is Brooke Struck, research director at TDL. And I’ll be your host for today’s discussion. My guest is Olivier Sibony, former partner and 25-year veteran at McKinsey, currently teaching at HEC Paris and author of the recently translated, You’re About to Make a Terrible Mistake. In today’s episode, we’ll be talking about nudging at the very top, not just nudging clients, not just nudging employees, nudges built by and for the C-suite executives that create strategy in large organizations. That’s the topic of Olivier’s latest book, and we’re thrilled to have him on the show. Olivier, thank you for joining us.

Olivier: Thanks for having me.

Brooke: So let’s get right down to business. Decisions, decisions, decisions. In your book, which I absolutely loved, by the way, you noted patterns of bad decisions, especially regarding big strategic decisions made by the leaders of organizations. Could you describe what types of decisions you are noticing and how did these fall short?

Olivier: There are many. In fact, I described nine in the book and I’m sure you could find others. And these are just from my observations. The background to this story really is that even as a beginner, as a young consultant, I was sometimes surprised and sometimes shocked that my clients who were these great, amazing people, who were these accomplished business people who were CEOs of large organizations or business unit heads, they would sometimes make decisions that really surprised me. I mean, and not because I was a genius. I certainly wasn’t. I certainly am not today but because it was obvious to anyone looking from the outside in that what they were doing did not make much sense. And when I would ask my colleagues about this, they would say, “Yeah, yeah. Sometimes clients are like that.” But we really didn’t have a very good explanation for why those things kept happening.

Olivier: So one example for instance, the one I started the book with is my very first client who was so keen on making an acquisition, a big acquisition in the US that when we essentially proved that it didn’t make sense economically, that the synergies would never justify the acquisition premium he had to pay, he basically told us, “Yeah, nevermind. I’ll just make a big bet on the currency, the big currency exchange bets and it will offset the loss that I will make in that acquisition.” And of course it’s flabbergasting to hear something like that from a CEO. And usually it’s not that explicit but when you’re keen on doing something, on making a big move, on doing something that will be remembered, on taking a big risk, you will find any justification for taking that big risk. That’s one example. There are many, many more.

Olivier: One, by the way, is the exact opposite example, which is to stare at something that is about to hit you and to say, no, it’s not a threat. That’s something that we’ve all seen in companies that are being disrupted by new technologies, by new competitors, by new business models and who keeps saying against all evidence, “No, it won’t happen to us. Digital disruption, we’ve seen that it happened in other industries but it can’t possibly happen to us. We’re different.” That’s the exact opposite pattern of the overconfidence of the big acquisition but it’s just as damaging. So there are quite a few of those and my interest in behavioral science and cognitive psychology came from basically trying to find an explanation, trying to find a logic for: why do these great people make decisions that are so surprisingly suboptimal?

Brooke: And you notice some patterns in there. You mentioned already that you found nine types. The fact that they’re types and not just scattershot already indicates that there’s some systematicity, there’s some logic behind what’s going on. They’re not random.

Olivier: They are not random. One of the reasons why a behavioral explanation is a satisfactory explanation for this pattern is that you would normally expect that people would learn, right? You would expect that when my firm, McKinsey and other firms at the same time keep writing articles about disruption and how it hits incumbents and how they are blind to it and how they don’t see the signals and how they respond too little too late, people would take notice. They would learn from the experience of others. You would hope that after hundreds, perhaps thousands of papers have been written about how acquirers tend to overpay for acquisitions because they tend to overestimate synergies, they would become more careful in their estimation of synergies, but no, they don’t.

Olivier: So how can we explain? And the rational model fails to explain that people don’t learn. The only satisfactory explanation for the fact that people, and again, great people, I mean, not losers, right? I mean, the best and the brightest keep making the same mistakes that they’ve made before and that other best and brightest people have made before them. The only explanation for this has got to be something that is a little bit beyond what we would normally call rational, although I try to use that word sparingly, but you see what I mean?

Brooke: So there’s clearly a cognitive dimension to that. Not that you’ve presented these overwhelming arguments. I’m a good receptor for that position, right? When you tell me that some cognitive dynamics are underplaying these things, I’m not inclined to challenge that situation.

Olivier: I realize I’m on my home turf here.

Brooke: Yeah, that’s right. But I want to push you in a slightly different direction. I think you catalog very nicely in the book, and then you’ve described already a little bit in our conversation today, some of those cognitive dynamics. But how is it that it’s not only individuals who fail to learn. It’s actually institutions that fail to learn as well. Is there something interesting to say about the cognitive dynamics of an institution as a thinking, acting organism?

Olivier: This is a fascinating topic and it’s actually a tricky one because organizations don’t have cognition. I mean, organizations don’t think. People think. And organizations operate in various ways. So we need to find a way to get from the individual level to the organizational level. We’ve got a level of analysis problem here. And my read of this is that organizations fail when they fail to overcome the cognitive biases of individuals. So the default setting, if you will, is going to be that if I’m an individual in an organization, I will tend to make my decisions to the best of my abilities and with my cognitive and other limitations. And the job of the organization, the task of the organization is to correct me if I’m wrong, basically. And most organizations in my experience don’t do that very well or rather they do it very well for some things and not very well for other things, which is why you tend to have recurring patterns of error.

Olivier: I’ll give you another example of one of those veterans, another one of the nine, which is the tendency to imitate successful models or successful so-called best practices. I can’t tell you how many times the client, when I was a consultant, has asked me, “How does that company solve this problem?” And that company was usually Apple or Google or Toyota or some company that was admired for whatever reason but which had nothing to do with the company in which the client who was asking me the question was sitting. And I learned quickly that the intellectually correct answer, which is “Why would you care? Let’s solve this problem for you,” is actually not a very productive answer because people do want to copy a solution that has been proven to work somewhere else. There are very good reasons for them as individuals, both emotional reasons and rational reasons, for them to want to do that.

Olivier: So an organization that works should be one that challenges you. When you come up with a solution and you say, “Hey, let’s do this because General Electric does it.” An organization should challenge you and ask you, “Well, great. But why would it work for us?” One example, speaking of General Electric, that has been seen in many organizations, especially in many Fortune 500 companies, is the fashion of forced ranking, which has now died down, thank God. But you remember how Jack Welch at GE was famous for implementing a forced ranking system where employees would be ranked in various categories with a mandated distribution, et cetera. And somehow people would associate the success of General Electric with that practice. And Jack Welch, by the way, encourages them to make that association and was a very vocal advocate of a forced ranking.

Olivier: Once companies started applying this, they realized that what worked for GE didn’t work for them, which really isn’t that surprising. Why would something work for you—if you’re a software company or a pharmaceutical company—why would it work for you just because it works for GE? That really doesn’t make much sense. But somehow the tendency to believe in best practices is so strong that literally hundreds of Fortune 500 companies adopted this practice and then of course realized it didn’t work and moved to something else. What a successful organization does, what a good organization does, is when someone comes up with a proposal like this, they actually challenge it a little bit. They have a decision-making process that doesn’t just let people come up with whatever idea seems right, which is what we all do as human beings. It actually has some checks and balances in the process and organizations fail when they fail to have the checks and balances that they should ideally put in place.

Brooke: So let’s talk about some of those checks and balances. So taking this GE example or you mentioned clients asking you, “Well, how does Google solve this problem or Apple solve this problem?” When I hear that, one thing that comes across, comes through my mind is no one ever got fired for trying the thing that Apple tried.

Olivier: Absolutely.

Brooke: There’s a risk mitigation there that even if the entire approach goes terribly wrong, the individual who proposed it is unlikely to be criticized for having put forward the idea that some wildly successful other company is doing. Whereas if you come up with something novel and you try that out, it’s you that’s on the line. You as an individual, as an employee, you take on a lot of risks by putting forward something novel.

Olivier: Absolutely.

Brooke: That’s something you talk about in your book quite a bit, and something I want to put a bead on and really clarify: the seeming schizophrenia that comes with this. So the example that you opened our discussion with (and that you opened the book with) about this massive gamble, this massive speculation on currency exchanges, seems like incredible risk-seeking behavior. But at the same time, organizations, large organizations, will also be incredibly risk averse. So how do we make sense of the fact that these large organizations are both incredibly risk seeking and incredibly risk averse at the same time?

Olivier: So it’s actually not as paradoxical as it seems for a couple of reasons. The first one is the one you’ve mentioned, Brooke, which is that people individually do not want to be the ones taking the risk. So you said nobody ever got fired for copying Apple or for buying IBM for whatever the safe choice is. And so individually, people will not be inclined to take risks. Dick Thaler, in Misbehaving, writes about this and actually explains that he doesn’t really see a solution to this problem. He calls it one of the toughest challenges that a CEO can struggle with in an organization to push people to take risks that in aggregate makes sense for the organization but that for them individually are too risky because no one wants to be associated with a risky gamble that failed. Even though at the time the decision was made, it may have been the right decision.

Olivier: In hindsight, people are smart enough to know that they will always be blamed for something that fails, even though it was a risk worth taking. So that explains part of the risk aversion. Another thing that explains the risk aversion but also the risk taking is a different dynamic, which is that organizations are usually very hierarchical, even when they don’t look hierarchical, even when they look cool and friendly and relaxed. They can be extraordinarily hierarchical when it comes to the big decisions and actually they value consensus. That’s an even stronger force than the hierarchy. It’s the groupthink that arises in organizations because people hate to disagree with one another. They need to work together and they need to like each other and divergence, dissensus is actually difficult to live with. People prefer consensus and that creates a pressure for groupthink.

Olivier: So when you put all those things together, organizations are designed for consensus around a hierarchically mandated form of risk aversion. Basically organizations are a machine to produce risk aversion under the control of the hierarchy and the benign control of the social group around the people who should be taking risks—which by the way is why we have a whole strain of management literature celebrating rebels, right? You’ve read those books that say, we should encourage rebels. And we should tell people to rebel against organizations. It’s interesting. I’m not saying it’s a bad idea. It’s probably a good idea to encourage people to rebel. But if you need to encourage people to rebel, you have a problem obviously. I mean, if the way your organization is designed is such that it can only work if people overcome the barriers that the organization puts in their way, well, maybe we should try to solve that problem rather than telling people to be rebels. Right?

Olivier: So basically an organization who will design the corporation because it’s well-designed is the machine that produces risk averse consensus. What are you going to do if you’re the CEO of that organization? If you are a reasonably successful, you are going to look at the money piling up in your bank account, and you’re going to find that you generate a lot of cash and you’re going to distribute some of that cash to your shareholders who are going to be very happy, but after a while you are going to realize that there are lots of startups out there that are creating businesses that are eating your lunch and that with all your organization and all your talents and all your people and all your brands and all your technologies and all your resources and all your distribution networks, you don’t manage to produce the innovation and the risk-taking that those little organizations out there are doing. What are you going to do in despair?

Olivier: You’re going to go look for one of those big bets. Like the one that my first client was making, and because you’re at the top of the organization and because you’re the CEO and because the organization is hierarchical and because it values consensus, that great idea of yours is not going to be subjected to the same checks and balances and vetting that every small innovation idea is subjected to. On the contrary, the hurdles are going to be much lower because it has to be done in secrecy because it’s a big bet because there aren’t that many opportunities because the bankers who are proposing it to you are telling you that you need to decide quickly, you know those dynamics, right?

Olivier: And so you have two completely different dynamics that are completely at odds with what you would rationally want to happen, where you would want to encourage a lot of small bets, because statistically, a lot of small bets that if each of them make sense in terms of expected value, that’s a very good idea in any organization, that is going to be actively discouraged. On the contrary, on the other end, you’re going to have a few very big bets sponsored at or very near the top of the organization that aren’t going to get the same level of scrutiny essentially. So that explains the paradox that organizations can be both very risk averse for small things and surprisingly risk-seeking for large ones.

Brooke: So what do better decision systems look like? How do we correct this inversion to get back to this ideal that you described—of having a proliferation of lots and lots and lots of small bets generally starting at the bottom and bubbling their way up (rather than nothing from the bottom managing to survive the trip to the top) but the things that start at the top ending up getting not nearly enough review and scrutiny and this sort of thing.

Olivier: It’s not easy obviously. But the inspiration is, well, the basic idea is it doesn’t happen by accident. It happens by process. You need to think about the processes by which your organization makes these decisions. And you need to customize those processes to encourage the small risks instead of discouraging them and bring a healthy dose of challenge to the big risks instead of letting them pass through simply because you’re the CEO. And those processes are going to be different processes. For the latter, it’s obviously a governance process. So you need to have a strong board that is going to be able to stand up to the CEO. That’s not easy. There is groupthink within boards as well. I tell some stories about that in the book. When it comes to the organization itself and the way you encourage the risk-taking, I think a lot of this boils down to how you actually prepare the proposals, run the meetings, discuss the numbers. It’s about the nitty-gritty of how those processes are designed.

Olivier: I think you can change a lot more than people believe by changing small things in the processes. And the sorts of things that you can change have to do with the dynamics of how the decisions are made, the quality of the dialogue that you have in the meetings where people make those decisions, the level of divergence of debate that you encourage between people in those meetings. And you can actually put in place practical tools and techniques that are going to encourage what I call dialogue, divergence and dynamics (the three D’s of good decision-making) in order to improve the quality of decisions systematically.

Olivier: The key idea here is (if you are the boss of, it could be a Fortune 500 company or it could be a team of five people, the problem is the same): If you are the boss, and you think you only need a decision-making process when you’re in doubt, when you’re not sure that you’re making the right decision, you’ve already got a problem. You need a good decision-making process to challenge you because you won’t know when you’re wrong. If you only put in place those dialogue processes and decision processes of all kinds when you’re not sure, you’ve probably made a lot of mistakes already, which by the way could be mistakes of taking too much risk or mistakes of not taking enough risk. They could go in either direction. But basically you want to put in place those processes in advance as safeguards against your own cognitive biases and against the biases of other people in the organization. But basically you need to start from the premise that only a good decision-making process in the organization can save you.

Brooke: So one of the things that I like about the way that you’ve framed that is that it gets at this thorny issue (that I’ve always found quite frustrating in my career) around corporate culture—that a lot of the discourse around corporate culture is like, “Well, corporate culture is both enormously powerful, it’s a huge determinant of outcomes. But at the same time, it’s completely ineffable.” And we wring our hands about how we wish we could influence corporate culture but it’s totally impossible. Culture is spontaneous. There’s nothing practical you can do. You just have the culture that you have.

Olivier: Completely amorphous. So it becomes a perfect excuse for everything that goes wrong or right. And a perfect explanation for everything that goes right.

Brooke: So I really like that what you’re proposing here is actually very concrete and tangible things that people can do to have an influence on the corporate culture and to achieve something closer to what it is that they want.

Olivier: When people tell me, “This will never work in our culture,” which is quite often the case, I tell them, “Look, I don’t know what will not work in your culture. Let’s try a few things. Let’s talk about how you encourage debate for instance.” So some companies I know have a formal process that looks like what I call a red team from the CIA’s Red Teams. So if you have a team that is recommending something, say a new investment, you appoint another team and you tell them, “Please show us how stupid that proposal is. Destroy the proposal of the first team, of the blue team.” And a lot of people tell me, “My God, this would never work in our organization. I mean, this level of conflict, of overt hostility, that would be completely unacceptable in our culture.”

Olivier: Fine. I totally get this. This is not for everyone. Right? Let’s try something else. Let’s ask for instance, this is another one of those techniques. Let’s mandate that every time someone comes up to you with a proposal, you ask them to give you two proposals, not just one. Okay. “Well that would work in our culture. We’re very open to that kind of debate.” Well, great. That’s just another way. It’s another tool. It’s another technique to reach the same goal, which is to open the aperture and not just to focus on one proposal. Why does it work? It’s not exactly the same thing. It doesn’t work in exactly the same way obviously but why does this particular idea work? Because suppose you have your three divisions reporting to you. And each of them is in the mode of trying to sell you the idea that they have in order to get the resources and the people and the cash and everything that they need to implement their idea and to make their division grow, which is understandably what they want to do.

Olivier: Now, suppose that division A has three great projects and division B has in fact zero great projects. Just by asking each of them to show you two instead of just one, you increase the chances that you are going to approve two of the best projects and not one of the worst projects, which happens to be proposed by another organization. You haven’t created conflict because this is a softer approach but you have improved the quality of your decision. You could have achieved the same goal by having someone challenge the content of what division B was proposing and showing that it was actually a bad idea but you don’t really need to do that if you think that is culturally unacceptable. What’s the point of this example? There are lots of ways. There are lots of techniques. There are lots of approaches to encourage dialogue and divergence and good decision dynamics. You just need to decide to put them in place once and for all and to apply them systematically.

Olivier: If you only say, “Well, I want to hear two proposals,” after you’ve heard division A and division B. And you think that division B’s proposal is not as good as division A’s then you’re starting to be a little bit unpleasant to division B. So you want to put in place those processes in advance, and you want to decide which ones you’re going to put in place. And they’re really easy to put in place. When people say, “It’s hard to change culture.” But when you actually show them that they can put in place meeting tools, decision protocols, design tools, and those very practical things, they realize that these are the levers by which you change culture. You don’t decide to change culture by using a magic wand. What you do is you say, “Here’s how we’re going to run our meetings.” And you set an example and gradually cultures do change.

Brooke: But I always thought that culture changes by printing the word “Innovation” in big letters on the wall, and then continuing to accept no risk that bubbles up from the bottom.

Olivier: Yes. “Innovation,” “customer focus,” and “ethics” are usually the three things that you put on the walls and look at how well it works.

Brooke: Yeah. One thing that you mentioned very, very briefly in passing in the last thing that you said—that I think is actually a real nugget of gold—is this idea: that what we should be striving for is a portfolio approach towards risk, and in implementing new techniques to change culture you suggested very briefly and just in passing that actually it’s the portfolio approach there too. There’s an entire playbook full of techniques out there to draw from the ideas, not that you know from the outset which of those techniques is going to be successful in your organization. And if we therefore think about the implementation of these techniques through the same lens as these innovation bets, well you take exactly the same approach. You take a portfolio approach, you see which things work and you test them out and you kill the ones that don’t.

Olivier: That’s absolutely right. And in fact, you test them out. You kill the ones that don’t work, which isn’t that often. More often, you actually add technique, upon technique, upon technique and you end up with a very refined decision-making process. Now, if I showed you what the decision-making process like this looks like, you would be horrified because you would say, “This is too complicated, it’s going to take time. It’s going to be bureaucratic, it’s going to slow us down.” And if you try to do all of those things overnight, that might actually be true because you’re not used to that. Now, if you start with one technique, I gave you the example of asking for two alternatives, two projects instead of one, take another one which is really easy to implement tomorrow morning in every meeting that you run.

Olivier: Before you do your little round-robin, where you go around the table and you ask people what they think, which you’re always going to do at some point in a meeting, at least I hope you’re always going to do it because if you don’t even listen to your people, well, then we’ve got a bigger problem. But suppose that you do actually listen to your people, just ask them to write down what they think before they speak. Take a minute. It’s only going to take one minute, right? Don’t tell me it wastes a lot of time. It’s only going to take one minute. Ask them to write down what they think of the proposal that they are going to be talking about before you go around the table. It’s an incredibly potent antidote to groupthink because once you’ve heard three people say, “Yes, boss, it’s a really great project that you’re proposing.” It gets just a little bit harder to say, “Well, I actually have some doubts.”

Olivier: So if you’ve actually taken a minute and written your doubts first, before you heard the three people speak, it might prod you just a little bit more to actually speak up and to speak your mind. That’s one example. It takes one minute. Now, that doesn’t change your organization, it doesn’t change your culture. In itself it’s not a very important change, but add another one like this and another one and another one gradually and soon you’ll have a completely different type of meeting and a completely different type of organization. And it will feel like a completely different place. So it’s a portfolio but it’s also a gradual approach where you constantly improve and you constantly add new tools to your decision-making process to make your decision making more and more professional.

Brooke: So having heard everything that you’ve said, and having read your book, I’m completely convinced that this is absolutely the thing that we should be doing. And as a completely rational agent, starting the moment that this call is over, I’m going to go and start adopting these techniques.

Olivier: Great.

Brooke: Now, of course making the rational arguments that these techniques work and that you can achieve better outcomes with them is only part of the battle, right? If we were talking to perfectly rational agents, we presumably wouldn’t be trying to overcome the types of challenges that we’re discussing right now. So how do we actually move the needle towards adoption now that we’ve identified some practices that actually do help knowing that what we’re talking about are boundedly rational individuals, they’re real institutions that have real cultures attached to them and have real decision-making processes in place. How do we move the needle on adoption beyond just the rational argument that it’s a smart thing to do?

Olivier: So a few things, first of all, we clean up our own mess before we start cleaning other people’s messes. So a lot of people who read my book or who attend my seminars say, “I’m going to go tell my boss that the way we make decisions on his team is not good.” I really do not recommend that you do that. You will not win friends and you will not influence people. This is not going to work and it might actually end your career faster than it should end. So start with your decisions. The focus here should be on how do I improve the quality of my decisions and my team’s decisions and the decisions that are under my control. It’s not about “you have biases”. It’s about, “I have biases and I want the decision-making process to be a safeguard, an antidote to my own biases”.

Olivier: So that’s an important question about where to start. And that means that if you are the CEO, well, it’s going to be on your board and your management team. If you’re the head of a business unit, it’s going to be on the management committee of the business unit. And if you’re the head of the sales team in Montreal, it’s going to be within the sales teams in Montreal. And that’s great, whatever the level where you do it, it’s going to improve the quality of your decisions. So first idea, do it in your sphere of influence. Don’t try to, and it’s an odd thing for me to say, right? But don’t try to influence the rest of the organization other than by your example, when people see that you succeed and that your team is energized and that the quality of your decisions and the quality of your results is improving, they should want to emulate that.

Olivier: The second practical piece of advice I would give to people who want to implement this is start small. Again, when people read the book and attend my classes, they sometimes say, “My God, there’s so much good stuff in here. Let’s do a dozen of those things.” And I really push hard to tell them, “No, please, please pick a maximum of three things. And once you’ve picked three things, pick the one of the three things that you are actually going to try tomorrow morning.” So if it is, write it down first, try that one. If it is the pre-mortem, which we haven’t talked about but it’s another tool that you are probably familiar with, try that one. Why? Because at first, almost any of those tools, even the most innocuous ones, is going to raise some eyebrows, is going to raise some red flags, is going to generate some resistance.

Olivier: Suddenly the boss is coming to me and asking me to write down what I think before I speak up, is he trying to trick me? What exactly is going on? Right? So you need to show that you mean well, and that the change you’re proposing is a change in your own behavior and not fake. You need to show the sincerity of your intentions. This is not another procedure. This is not another form that you’re asking people to fill, right? When we say you.

We want to hear from you! If you are enjoying these podcasts, please let us know. Email our editor with your comments, suggestions, recommendations, and thoughts about the discussion.