Since first articulating his revelatory insight into how incumbent firms fall prey to disruption, Clay Christensen of the Harvard Business School has continued to work at the ambitious task of “theory building.” The realm he is trying to create models to explain – the effective management of innovation – could not be more central to business success. Managers today know that their enterprises depend on frequent, important innovation, but they lack good frameworks and tools to act on that recognition. Management science as it is taught today and embedded in firms’ structures and processes still assumes that the introduction of a new offering – let alone a new business model – is the exceptional event and not the norm.
On the eve of the 6th Annual Global Drucker Forum in Vienna, where Professor Christensen will provide a keynote address, I asked for a preview of the territory he will cover. The following is a lightly edited transcript of our conversation.
HBR: There have been concerns expressed lately that, in advanced economies, we’ve thoroughly exploited the major technological breakthroughs of the past century and might now have hit a plateau. Further gains in prosperity will be very hard won. But some of us are hopeful there are still huge leaps to be made in the management of large organizations. What do you think: do managers matter? What would make their work matter more?
Christensen: I think that the evidence is that managers themselves have resigned themselves to not matter. There are so many practices that they are engaging in that cause management not to matter.
But your questions point directly to why Peter Drucker’s work is inspiring. I know that I, from an academic point of view, want to make a difference to managers. I truly do. And I know a lot of other people who truly do want to make a difference to managers. There was something about Peter Drucker that enabled him to do that even though, to my knowledge, he never explicitly said “I want to influence the way managers think.” Can we learn anything about the way he went about his work so that he made a huge difference almost in spite of himself?
I have a hypothesis about what he did. For a book I’m just finishing I have written an essay about what theory is and how it is built. I describe three stages for building a theory. The first stage entails essentially dumpster diving – you actually have to just dive in and hang out with the phenomena to try and figure out what is going on in the world, and you actually need to create data. Note that data actually is created by mankind. God didn’t come down and hand us data, and Mother Nature has not taken her data and input it into a balanced data set for our convenience. We actually create data, and the big question is: what data do you create? The answer makes a huge difference in what you try to optimize.
Then in the second stage, which we’ll call the empirical theory stage, you manipulate the data you’ve created to find out what it has to say. You are always aware that data isn’t “real” – it is just a proxy for the phenomena. But the nice thing about data is that it allows you to move yourself to a level of abstraction from the phenomenon and get a sense for how things work. Having a proxy means you can do tests: you can subtract this from that, regress this versus that, chart this or that.
The problem is that most academics, in their attempts to create theories, begin and end at the level of data. And increasingly they don’t know what is going on in the dumpster because they haven’t spent any time living there. They didn’t create the data – so they manipulate data without even recognizing that it is a proxy for reality. They think it is real.
You’re reminding me of that great admonition that “the map is not the territory.” Abstractions always limit your ability to see the full reality.
But it’s a great thing to have data at an abstract level, because you can manipulate it – whereas you can’t manipulate the people. You just have to keep the data in perspective.
And when you’re done with that second stage, it’s time to move to what we’ll call a complete theory. Again using the data, what you are trying to figure out is what in academia we call the construct. If you have a construct, then it’s just a hop, skip, and a jump away from understanding what really causes things to happen.
What would be an example of a good construct in management?
For example, Joe Bower wrote this marvelous book called The Resource Allocation Process in which he asked: As ideas come up from the bottom of the company, how do the middle managers decide which of these ideas will they push forward to the senior people who will then invest? To answer that, he gave us a construct he calls “impetus.” Some ideas get impetus and come to the top, and other ideas don’t get impetus and they languish. Once you are exposed to that construct – my gosh, your understanding of how the resource allocation process works is so far advanced.
In my research, I look at the trajectories of technological improvement and the trajectory of customers to utilize improvement – and those are constructs. And the difficulty of getting those right is part of why I admire Drucker’s accomplishment. Because in essence, that is what Peter Drucker did for mankind: in every element of a manager’s life, he gave us a construct that allows us to think about things differently.
The “dumpster diving” you’re describing seems like it was dear to Drucker’s heart. I read recently that he enjoyed close friendships with economists Joseph Schumpeter and Karl Polanyi – but ultimately chose not to become an economist himself because he found flesh-and-blood people more interesting than abstract models.
Yes, yes. And I’m just thinking out loud here, but I have a very close friend named Gordon Williams who is a professor at Harvard Medical School. And he is probably the best in the world at endocrinology and hypertension. When I interviewed him about why he has been so successful, he said that almost everybody else in the business and in academia works at the level of data. They’ll compare sets of patients with this outcome to patients with that outcome, and determine if you give them this drug, what is the probability that it will or won’t work. But at its core, he said, a disease is a process.
That is actually very important. A disease is a process, and you cannot understand the process that constitutes a disease if you have an “n” of 10,000 patients. At best you can correlate characteristics of patients with outcomes of interest. It is only at an n of one that you can understand deeply the processes inside the patient. In the end of course you are not trying to understand the patient; you are trying to understand the disease. You want the ability, when a patient comes in, to very quickly say, “you don’t have that disease; you have this disease.” But in order to understand the disease as process, you have to study an n of one.
By analogy we can look at Peter Drucker’s methods. Corporations actually have diseases and they are rooted in processes inside of companies. I don’t think Peter spent a lot of his time manipulating data. He tried to understand processes that go on inside the company so that he could understand, “you have this disease, and you don’t have that disease.” Of all the contributions that Peter Drucker gave to managers, above everything else is that he taught us how to think.
In your most recent article for HBR, The Capitalist’s Dilemma, you make a brief reference to Drucker’s work, and also Ted Levitt’s, and you write that the practice of management has “regressed” since they made their contributions. Do you really believe that?
I do. I do. Where this conviction comes from is from my doing deep dives in the dumpster. I’ve been crawling inside companies and listening to what managers talk about. What is the agenda when they have a management meeting? A shocking proportion of their conversation is spent on topics like “where are we going to get the numbers so that we can deliver the numbers to Wall Street on time?”
My observation of what’s happened is that, today, the people convened for that senior executive meeting all have different languages that they speak. One speaks finance, one speaks HR, one speaks operations, and so on. They have discovered that, if they translate all of their initiatives into numbers, then everyone can talk about them.
So numbers are the lingua franca of the C suite, but that forces everyone into a very limited vocabulary.
Yes – when you put the agenda together, all of the options that people need to decide upon are translated into numbers. And so the evaluation of the ideas quickly turns into a review of how good the numbers look, as opposed to being a substantive discussion about things that are not known. That’s why I think we’ve regressed.
And this brings us back to: Do managers matter? They actually don’t matter unless they are trying to get ahead of themselves and create the insight to frame what is really going on in the world. And to do that, you actually have to interact richly with all of these people and all of these problems, and then develop data to describe accurately what’s going on. If you are not going in and out of the dumpster, you can’t do your job.
One important notion in the Drucker Forum this year is this idea of having arrived at a turning point. In what way, if any, do you think we have?
We are at a fork in the road and it’s a big fork in the road. Much of me believes that the role of finance in our economies will diminish very significantly in the next ten years. Why do I say that? In any industry, if the product or service you are making is a commodity and is undifferentiable versus the competition, then you lose power. We noted in “The Capitalist’s Dilemma” that we are awash in capital. A few weeks ago there was an interesting article in The Wall Street Journal describing how if you have euros and you want to put them in a big bank, you now have to pay them to hold your money. The cost of capital is negative. Seeing these kinds of things happen, I think banks, many of them, won’t exist ten years from now. Their functionality will be taken over by IT companies who don’t have the same assets and income statement challenges.
If that is true, then the fork in the road for managers is this: if finance, which has been the kingpin for the last fifty years, is no longer the kingpin, what will be? The only viable alternative is talent, so that managers must learn better how to help people become more capable.
In one of Drucker’s last books he argued that in developed countries the productivity of knowledge workers would have to increase very rapidly or otherwise those countries and their organizations would lose position and become steadily poorer. He said “the most valuable asset of a 21st century institution will be its knowledge workers and their productivity.”
That is exactly right. There is no other lever that managers can pull or tweak or open or close that will have a dramatic influence on their future. And managers should welcome that, because that makes management fun again.
This post is part of a series leading up to the 2014 Global Drucker Forum, taking place November 13-14 in Vienna, Austria. See the rest of the series here.