In this edition of Author Talks, McKinsey Global Publishing’s Rama Ramaswami chats with Roger Martin, a thought leader in business management and former dean of the Rotman School of Management at the University of Toronto, about his new book, A New Way to Think: Your Guide to Superior Management Effectiveness (Harvard Business Review Press, May 2022). A cornerstone of business education is the idea that all decisions must be based on rigorous data analysis, but Martin says this framework is beginning to crumble under the weight of an uncertain future. Some traditional management models assume that future trends will mirror past ones, but in a rapidly changing world, this certainty is becoming less and less reliable. An edited version of the conversation follows.
Rethinking management models
Why do we need to think differently about the essentials of management?
A New Way to Think is about how you can take more ownership of the models you use in order to be a more effective manager. This is a return to the mainstream of what I’ve always done, which is to write about management. I’ve become more and more interested over time in how many times a model becomes the accepted wisdom—regarded as “this is the way we think about this problem”—how many times it becomes so entrenched, so much so that even if it doesn’t work and has a track record of not working, we still stick with it.
What I wanted to do is write a book that hopefully made managers brave enough to throw off some of the models that don’t work for them. It was an extension of my work on how managers think, and the patterns of thinking that guide their actions.
What is wrong with traditional frameworks, many of which have been around for decades?
A number of them are crumbling under their own weight. For example, there is the idea that in order to maximize shareholder value you should provide your senior executives—and your CEO in particular—lots of stock-based compensation incentives. That’s been around since the late ’70s. It really accelerated, and I would say became the dominate model, by probably 1990.
You’d think you’d see a market positive difference for people to keep doubling down on that model, but there’s no evidence that it’s worked for shareholders. You could argue that a bunch of things that are bad have happened since to get the stock price up so CEOs get compensated more. That’s been the product of the model. Isn’t it time we asked the question: Is there a better way to do that? I would say that’s happened across a number of models.
Our model coming out of business education is that all decisions must be based on rigorous, fact-based analysis, and if they’re not, it means you’re not paying attention, you’re lackadaisical, you’re slack, you’re old fashioned. In fact, the assumption under analysis is that the future must be identical to the past—otherwise, the analysis you did is based on a sample that isn’t representative of the universe, and you were taught in Statistics 101 that that’s a bad idea. If you went to business school or if you went to engineering school, you were taught that all decisions need to be made on the basis of fact-based, rigorous data analysis.
I would say we’re doubling down on models that don’t have the outcomes that we wish they did, and it is time to take a look at those models.
Our world is so unsettled these days. How can we design a strategy that works both now and in the future?
I think it’s a myth that you can design a strategy that will for sure work in the future. Last time I checked, nobody knows the future—maybe some higher power—but none of us regular humans know the future. It unfolds as it will. So all strategy can do is improve your odds or shorten your odds. You have to take into account the fact that things may evolve in a way that’s much different from what you expected, in which case you’re going to have to look at your strategy and ask, “How do I have to tweak or modify it?”
My first view would be that you’re setting completely unrealistic goals if your goal is to be certain that your strategy can be stable and work for the future. The other thing I’d say about strategies for the future is that, again, if you base your strategy on analyzing the past, then you are implicitly making the assumption that the future will be identical to the past. Because it’s based on rigorous analysis—and you’ve been taught in business school that rigorous analysis is correct—then you will not be ready for the future to end up looking different than the past, and you’re more likely to stick with your strategy for longer because it’s “right,” based on the analysis. As a result, you’ll stick with the strategy longer and probably crash worse.
If you base your strategy on analyzing the past, then you are implicitly making the assumption that the future will be identical to the past. Because it’s based on rigorous analysis—and you’ve been taught in business school that rigorous analysis is correct—then you will not be ready for the future to end up looking different than the past, and you’re more likely to stick with your strategy for longer because it’s ‘right,’ based on the analysis.
If instead you take strategy as being an exercise in dealing with the future, then you will take Aristotle’s advice. Many people are not taught this and don’t realize this, but he invented the scientific method. It was formalized by Bacon, Newton, Descartes, and Galileo 2,000 years later during the Scientific Revolution, but he laid it out.
The father of science, Aristotle, warned that you shouldn’t use this methodology when you’re dealing with a part of the world where things can be other than they are. He said his methodology was strictly for one part of the world, the part of the world where things cannot be other than they are. Aristotle said in that part of the world, you can analyze the past and be entirely certain that whatever understanding you come to, you can be entirely confident that it will happen in the future.
He said there’s another part of the world where the future can be different than the past. This is like the digital world, for example. In 1999, there were zero smartphones—zero. The first was the BlackBerry in 2002. The last time I checked, there were 4.4 billion smartphones in the world—that’s the part of the world where things can be other than they are. What Aristotle said is, in that part of the world, never, ever, ever, ever use the scientific method. Isn’t that interesting?
What he said is, you must imagine possibilities of that future and choose the one for which the most compelling argument can be made. So I think strategies will be more robust for the future when we stop imagining that we can analyze the past to determine our strategy and rather understand that it is an exercise in imagination—imagining a future and then testing out the logic to ask which possibility for the future is the most logically compelling.
Notice that it’s about being logically compelling, not data based. It’s “here’s what would have to be true about our customers, here’s what would have to be true about our competitors, here’s what would have to be true about our capabilities.” You lay out what would have to be true and ask which condition seems most compelling. If you do that, you can do what Aristotle said. I think that’s the best way to deal with the uncertainty of the future—to shape it, not let it happen to you—but you have to have this imagination and a belief that you can shape the future rather than a belief that you can extrapolate the past. That’ll make for better strategy.
Less data, more creativity
You argue that making great choices requires more imagination than data. How do we put this into practice?
It’s mainly an exercise in freeing people up, in saying to them that in order to shape your strategy for the future, you don’t have to go by what the data say. What you can do is imagine things that you think would be better for us, better for the consumer, better for the world, etcetera, and then we’ll test those.
I think the world of business has put people in a straightjacket of strict, rigorous data analysis. Once freed from it, they are game to play. That having been said, when I work with companies and groups on that, I do encourage playfulness. If the ideas coming forward aren’t crazy enough, I say, “I’m not stopping until I get at least one idea that all of you think is crazy. If everybody doesn’t think it’s crazy, then it doesn’t count. We’re going to keep going until we get one of those.”
People are born with the ability to imagine; I think kids with their drawings can imagine anything. And then, policies and practices of organizations that people are involved with keep on hemming that in. This is why the book is about models. If the model is, “I am only going to believe it if it’s backed with rigorous, statistically significant analytical data,” I will just tell you sorry, that doesn’t cut it.
There are many, many executives who are now practiced in the art of backward fitting data to the imagination they’ve set. They’ll come up with an idea and say, “I want to do this thing,” and then they’ll tell their people, “Go put together a data pack that would support this.” I’ve seen it, I understand why—because they need a legitimizing factor to do the thing that they really wanted to do in the first place, for which they cannot point to rigorous data. Is it that they don’t have any data at all? No. They’ll say, “Well, I saw this in this other industry, and I was talking to this set of consumers the other day on this, and it occurred to me that we could combine those two things into this.”
So it’s not illogical. Those were logical leaps. It’s something that Charles Sanders Peirce, this great American philosopher, called abductive logic, which he called a logical leap of the mind. People are capable of that—they’re generally just taught that it’s not legitimate. Now, who is it legitimate for? Entrepreneurs. If I’m an entrepreneur and I say, “Here’s my great idea,” and you say, “That’s crazy. What proof do you have of that?” What’s the entrepreneur inclined to say? They’re inclined to say, “I don’t care that you think it’s crazy. This is going to work, and just watch me.” There’s this entire class of businesspeople—entrepreneurs—who ignore those models, and they’re the people who create the future by saying, “I’m going to make something right that isn’t.”
For example, Steve Jobs made a little white MP3 player with a little wheel on it and sold it for three times the going price of a similarly featured MP3 player that had been proven to be a solid piece of equipment. He soon had and kept over 50 percent market share of that business, at three times the price, and it’s because he made something true that wasn’t. What he said was, “Those little MP3 players would be way more valuable if we could create a really easy way to download songs from the internet onto that, one song at a time rather than an album at a time.”
Did that exist? No. Did he make it exist? Yes. Did that make him famous? Yes. Did he dent the universe? Yes, which is what Aristotle said: to be the cause of the effect you want to see, to dent the universe. That is logical, but it would be what Peirce called abductive logic, not the inductive or deductive logic that we are taught in business schools.
What now?
Millions of workers are quitting. Is there any right way to manage talent?
I think there is, but I would go first to quitting and the Great Resignation. I deal with this in a chapter of the book on the importance of habit. What I think these companies have not figured out is that habit is a huge driver of behavior. We think it’s conscious loyalty, something that you consciously think about, but it turns out that the conscious is the tip of the iceberg, and the unconscious is like the 95 percent that’s below the water.
The mind loves comfort—the comfort that turns into automaticity, so that you don’t even think about something. We were automatically in a habit of commuting to the airport or to your downtown office every day. You didn’t actually think a lot about it because it was just your habit—that’s the habit which you stop thinking about, so you don’t question it automatically. In fact, if you start to question that, your subconscious starts saying, “Wait a minute. Wait a minute. That’s our habit. What are you messing with? I’m comfortable with that habit. I may not love it, but I’m comfortable with it.”
What I think these companies have not figured out is that habit is a huge driver of behavior. We think it’s conscious loyalty, something that you consciously think about, but it turns out that the conscious is the tip of the iceberg, and the unconscious is like the 95 percent that’s below the water.
Then COVID-19 comes along, and what does it do? It’s a forced break in the habit. In fact, I would argue it’s the biggest forced break in habits since, for sure, World War II. We had this big habit break where you can’t go to the office, and you can’t go to your clients’ offices. You adopted a new habit, which is, I would argue, the habit that you’re engaging in right now, which is working from home. That was called working remotely. What I would argue is there’s a new definition of remote, and it’s the office—that’s now remote work.
That jarring effect is a big one. What does it do? It causes it to go from the subconscious of habit to the conscious of “I have to adopt an entirely new habit of working remotely, and that’s going to cause me to reevaluate the entire package.” Habit has now been broken again. Companies that are ordering you back to work are breaking a habit, and when you break a habit, you end up at the starting line with all other potential new habits.
Upon hearing that they’re going back to the office, all these people said, “Whoa, whoa, whoa, that’s a new job. That’s a new thing. I’m going to evaluate it against all other possible jobs.” They’re saying, “I didn’t really like my boss” or “I hate that commute” or all the other reasons that would have been suppressed by the subconscious before as part of the habit. So people are quitting left, right, and center, and it’s no surprise whatsoever.
That’s part of the question. The other part of the question is, how do you have to treat talent? The assumption in the world of talent is that talent is motivated by being paid the most. I do not believe that for one moment, because that’s not how talent operates. Talent is motivated by being paid fairly, but they’re motivated by an organization that treats them as a special talent, that recognizes their individual talent.
They don’t have to actually say, “Your individual talent is better than that person’s.” It’s just that they recognize you for the distinct package that you are. That’s so important to talent because to get to be high-end talent, the person has had to work on himself or herself all through their lives in education, work, and experiences. If they’re an athlete, it’s staying after practice to shoot more free throws or take more swings at the bat.
There are all these things that are unique to high-end talent, that made them who they are, so to be treated generically is something that grates on talent. The real secret to talent is to treat them as an individual, to make sure that they’re not blocked in their own individual path forward, and to make sure they’re appreciated and patted on the back for their individual ability to contribute.
There are all these things that are unique to high-end talent, that made them who they are, so to be treated generically is something that grates on talent. The real secret to talent is to treat them as an individual, to make sure that they’re not blocked in their own individual path forward, and to make sure they’re appreciated and patted on the back for their individual ability to contribute.
I don’t think that’s understood. Lots of talent shifts around and is viewed as going to the highest bidder, when that’s not what they’re doing. They’re going to a place that will make them feel special for their talent.
What surprised you the most when you were researching and writing the book?
It felt as though—if you took models to be animate things—models come to own their users. That’s what I came to believe that people need to do. People need to take account of and document what they expect from a model, because human beings have an infinite capacity to ex post rationalize. For example, a company may think a new investment is going to get a 50 percent sales increase within five years. Five years later, revenue increases 25 percent rather than 50 percent, and everybody will be nodding, saying, “Yeah, yeah, yeah. This is exactly what we expected—25 percent sales increase.”
What they have to do is write down 50 percent, put it in the drawer if necessary, and then pull it out and say, “Oh, it’s 25 percent, not 50 percent. What didn’t work the way we thought? Are we going to build another factory just like that, expecting 50 percent? No, we’re not.” My hope for people is that I give them encouragement to watch out for getting owned by their models, and to be more aggressive than they are now about owning their models, which means dispensing with ones that just don’t work.
What are some practical steps that companies and leaders can take to break away from old models?
One would be what I’ve just described, which would be recognizing when you’re using a model. People might not say they’re using a model, but they are when, for example, they start a meeting by going around the table and having people introduce themselves. That’s a model. Or they start a meeting by not doing that and instead yelling and shouting at people. That’s a model, too. People have to look at the actions and the thinking that’s guiding the actions.
If a boss is inclined, whenever he or she is unhappy, to start the meeting by pounding the table—that’s a model. The person in question who’s using that model should say, “Here are my expectations of how that’ll work: it’ll get people to snap to attention, get people to recognize that I’m really serious about this, and I will get better work from the people in that room.” It doesn’t work the way it was expected. Then, I would say you’ve given yourself at least the data, the insights, to be able to say, “Maybe I should try something else. Maybe I should try saying more calmly, ‘I’m really disappointed in our results and I would love to have us brainstorm our diagnosis of that. Then we can work together on a solution.’”
That’s a different model. That’s not pounding the table, that’s showing some vulnerability and asking for help. Let’s see if that model works better. The little steps to take are to make sure you are making the expectations clear to yourself of the model so that you can evaluate the results—so that you can experiment with others if your results aren’t what you wish they were.
Watch the full interview
