McKinsey Quarterly

Strategy’s strategist: An interview with Richard Rumelt

| Interview

As a mountaineer Richard Rumelt, a professor of strategy at UCLA’s Anderson School of Management, has achieved a number of first ascents. The same holds true in Rumelt’s academic career. In 1972 he became the first person to uncover a statistical link between corporate strategy and profitability, finding that moderately diversified companies outperform more diversified ones—a discovery that has held up after more than 30 years of research. Rumelt also challenged the dominant thinking with his controversial 1991 paper, “How much does industry matter?” His study, published in the Strategic Management Journal, showed that neither industries nor corporate ownership can explain the lion’s share of the differences in profitability among business units. Being in the right industry does matter, but being good at what you do matters a lot more, no matter what industry you’re in. This study was one of the first entries in what has since become a large body of academic literature on the resource-based view of strategy.

Rumelt holds the Harry and Elsa Kunin Chair in Business and Society at the Anderson School. Recently, he met in San Francisco with McKinsey director Lenny Mendonca and Dan Lovallo, a professor of strategy at the University of Western Australia.

The Quarterly: Richard, you’ve been teaching about, researching, and consulting on business and corporate strategy for 35 years. What changes have you seen in that time?

Richard Rumelt: Some of the biggest changes have been in the process of generating business strategies—what I call “strategy work.” Around 1980 the received wisdom was to decentralize into business units, which would each generate a strategic plan. These plans were then amalgamated up the hierarchy, in some portfolio way, for senior management. That approach has all but disappeared, and we’ve seen a dramatic recentralization of strategy work.

The Quarterly: Last year the Quarterly’s survey on strategic planning found an enormous amount of dissatisfaction among executives. Many of them feel that they are wasting a lot of time on strategic planning. What advice would you give them?

Richard Rumelt: Most corporate strategic plans have little to do with strategy. They are simply three-year or five-year rolling resource budgets and some sort of market share projection. Calling this strategic planning creates false expectations that the exercise will somehow produce a coherent strategy.

Look, plans are essential management tools. Take, for example, a rapidly growing retail chain, which needs a plan to guide property acquisition, construction, training, et cetera. This plan coordinates the deployment of resources—but it’s not strategy. These resource budgets simply cannot deliver what senior managers want: a pathway to substantially higher performance.

There are only two ways to get that. One, you can invent your way to success. Unfortunately, you can’t count on that. The second path is to exploit some change in your environment—in technology, consumer tastes, laws, resource prices, or competitive behavior—and ride that change with quickness and skill. This second path is how most successful companies make it. Changes, however, don’t come along in nice annual packages, so the need for strategy work is episodic, not necessarily annual.

Now, lots of people think the solution to the strategic-planning problem is to inject more strategy into the annual process. But I disagree. I think the annual rolling resource budget should be separate from strategy work. So my basic recommendation is to do two things: avoid the label “strategic plan”—call those budgets “long-term resource plans”—and start a separate, nonannual, opportunity-driven process for strategy work.

The Quarterly: So strategy starts with identifying changes?

Richard Rumelt: Right. Let’s take an example. Right now, the advent of 3G1 cellular technology makes it possible to deliver streaming video over mobile phones. Cell phone makers, cellular carriers, and media companies all need to develop strategies for exploiting this change. Even though these changes have long-term consequences, companies need to take a position now. By “take a position” I mean invest in resources that will be made more valuable by the changes that are happening.

For example, I think high-bandwidth opportunities are being overhyped in this 3G game. As Clayton Christensen2 has pointed out, technologists often overshoot consumer demand. I tend to think this is happening in the 3G arena, so I am much less interested in the higher-bandwidth applications, like streaming video, than in lower-bandwidth opportunities, like streaming audio and mobile search. Give me a cell phone that combines voice recognition with location-filtered search results and you have a product that a wireless company can differentiate.

Now, speculative judgments like these are the essence of strategic thinking, and they can be the starting points for taking a position. Can you predict clearly which positions will pay off? Not easily. If we could actually calculate the financial implications of such choices, we wouldn’t have to think strategically; we would just run spreadsheets. Strategic thinking is essentially a substitute for having clear connections between the positions we take and their economic outcomes.

Strategic thinking helps us take positions in a world that is confusing and uncertain. You can’t get rid of ambiguity and uncertainty—they are the flip side of opportunity. If you want certainty and clarity, wait for others to take a position and see how they do. Then you’ll know what works, but it will be too late to profit from the knowledge.

The Quarterly: So how does a company take a good position?

Richard Rumelt: Well, one big factor is a predatory posture focused on going after changes. Back in the mid-1990s I was researching strategy in the global electronics industry. I interviewed 20 to 30 executives, CEOs, and division managers and asked fairly simple questions. Which company was the leader in their market? How did that company become the leader? What’s their own company’s strategy?

I saw an interesting pattern. Most executives easily explained how companies became market leaders: some sort of window of opportunity opened, and the leader was the company that was the first to successfully jump through that window. Not exactly the first mover but the first to get it right.

But when I asked these same executives about their own strategies, I heard a lot about doorknob polishing. They were doing 360-degree feedback, forming alliances, outsourcing, cutting costs, and so on. None of them even mentioned taking a good position quickly when the industry changes.

Then in 1998 I had the chance to talk with Steve Jobs after he’d come back and turned Apple around. I was there to help Telecom Italia try to do a deal with Apple, but after that business was completed I couldn’t help asking a question. “Steve,” I said, “this turnaround at Apple has been impressive. But everything we know about the personal-computer business says that Apple will always have a small niche position. The network externalities are just too strong to upset the de facto “Wintel”3 standard. So what are you trying to do? What’s the longer-term strategy?"

He didn’t agree or disagree with my assessment of the market. He just smiled and said, “I am going to wait for the next big thing.”

Jobs didn’t give me a doorknob-polishing answer. He didn’t say, “We’re cutting costs and we’re making alliances.” He was waiting until the right moment for that predatory leap, which for him was Pixar and then, in an even bigger way, the iPod. That very predatory approach of leaping through the window of opportunity and staying focused on those big wins—not on maintenance activities—is what distinguishes a real entrepreneurial strategy.

The Quarterly: So he spotted—and then exploited—a change whose time had come?

Richard Rumelt: Yes, which isn’t to say the changes had been invisible. Lots of people in and out of the industry knew about music downloading—you couldn’t pick up a magazine without reading about Napster.4 And people knew that MP3 players were coming along. As in most times of change, you had major actors, with key resources, that didn’t want to act—in this case, the music companies and the music retailers.

Enter Jobs. He was perfectly positioned because he was a bit of an insider in the entertainment industry but didn’t have any of those asset positions that were being threatened. He didn’t need to make a fantastic leap of imagination into the far future. He found a set of ideas that needed to be quickly and decisively acted upon.

The Quarterly: What capabilities do companies need in order to take advantage of these ideas?

Richard Rumelt: There is no substitute for entrepreneurial insight, but almost all innovation flows from the unexpected combination of two or more things, so companies need access to and, in some cases, control over the right knowledge and skill pools.

Right now I’m following a little company called Sherline Products that makes machine tools for model makers. These are small machine tools you can buy for about $3,000, such as computer-controlled lathes and minivertical mills. Sherline sells them to model makers and to companies creating prototypes. Sherline’s CEO says he wouldn’t have been able to conceive these products if he hadn’t been both a hobbyist and a professional machinist. The professional machinist side of him knew what capabilities the machine tools really ought to have, and the hobbyist knew about operating in a small space with a limited budget. So he simultaneously had knowledge of two things that aren’t typically combined. That allowed him to create this product.

Similarly, the iPod came from knowledge and resources being adroitly combined. There were lots of people who knew the music industry and lots who knew about hardware and lots who knew about the Web. But to quickly and skillfully access those three pools of resources and knowledge was an impressive feat.

The Quarterly: So how do we know which changes are important and which resources to combine?

Richard Rumelt: That’s a very tough question. It is a key issue—the next frontier. And it is underresearched, underwritten about, and underunderstood. I call it “strategy dynamics.”

Most of the strategy concepts in use today are static. They explain the stability and sustainability of competitive advantages. Strategy concepts like core competencies, experience curves, market share, entry barriers, scale, corporate culture, and even the idea of “superior resources” are essentially static, telling us why a particular position is defensible—why it holds the high ground.

If the terrain never changed, that would be the end of the story. High ground is always high, and low ground is always low. But in business, unlike geology, change happens in years rather than millennia. In the modern business world, there are earthquakes all the time that quickly take the low ground and raise it high and, at the same time, submerge some mountain peaks below water.

Strategy dynamics studies how those changes would shift each dimension of an industry. Would the industry become more concentrated or less? More integrated or less? Would there be more product differentiation or less? More segmentation or less? Given consumer desires and available technologies, how should the industry or business look in, say, ten years? Where are the economic forces trying to take you? Should your strategy ride those forces or fight them?

There are tools and exercises that help trigger inductive insights about dynamics. One is a list of common biases—the kind of list that helps some people look beyond the standard consensus view of what is happening. For example, most analysts overestimate the importance of scale and underestimate the inertia of buyers, so what happens if we adjust our views to control these predictable biases?

Another useful exercise is to rethink the metaphor. During the telecom boom of 1997 to 2000, people were saying that fiber optic cable was like the microprocessor: capacity was rising exponentially while costs were fixed (as in Moore’s law.5 ) And just as the microprocessor revolutionized the computer industry, optical fiber would totally restructure telecom. But the metaphor was specious. Excess cable capacity has a very different effect than excess PC capability. Because network capacity is a shared resource, excess capacity can slam prices down to variable cost, which is virtually zero. By contrast, overpowered PCs have no real effect on price, because there is no market for using excess CPU cycles or excess memory. Once you see the underlying metaphor, you can adjust your expectations.

The Quarterly: What’s another way to understand strategy dynamics?

Richard Rumelt: I use another tool I call “value denials.” These are products or services that are both desired and feasible but are not being supplied to the market. The concept combines insights into demand and potential supply. A classic example is an airline ticket guaranteeing that your luggage will not be lost. It just isn’t supplied at any price. There must be a price at which airlines would hand-carry luggage to the baggage compartment and even a price at which they would strap it into the seat next to you! There are times when we would pay the premium, but those services are not offered. That’s a value denial.

A value denial is a business opportunity. Every change and innovation creates new value denials. People wanted to buy music à la carte and keep 10,000 songs on their computers. Well, they got that, but there was a value denial: the digital music wasn’t portable. So along come the MP3 player and the iPod. But those innovations uncovered a new value denial: people also want to plug their players into their stereos. Well, this was pretty easily fixed, but playing your MP3s on your stereo uncovers yet another value denial: MP3s are compressed and just don’t sound as good as CDs. Finally, even when I have immediate access to all music anywhere and anytime through the “jukebox in the sky,” there will still be a value denial—how will I know what to listen to? I will need a private tutor and disc jockey to help arrange my listening and maybe to shape my tastes.

So one useful way to think about change is to turn aside from the central innovation and ask yourself what value denials it will uncover. How will they be fixed? And what value denials will then be uncovered by that fix?

The Quarterly: What sort of group can analyze these kinds of things?

Richard Rumelt: A small group of smart people. What else can I say? Doing this kind of work is hard. A strategic insight is essentially the solution to a puzzle. Puzzles are solved by individuals or very tight-knit teams. For that, you need a small group. With big groups and complex processes you can select the better solution to the puzzle, and you can get consensus and buy-in and even commitment.

One other thing. If I had my way, small groups like this would be absolutely prohibited from doing PowerPoint presentations! Using bullet points so much drives out thinking. One of the nice features of PowerPoint is how fast you can create a presentation. But that’s the trouble. People end up with bullet points that contradict one another, and no one notices! It is simply amazing.

If you ask a group to put aside the bullet points and just write three coherent paragraphs about what is changing in an industry and why, the difference is incredible. Having to link your thoughts, giving reasons and qualifications, makes you a more careful thinker—and a better communicator.

The Quarterly: Shifting gears a bit, Richard, can you tell us about your research on diversification and focus?

Richard Rumelt: Well, my first research on corporate strategy showed that somewhat diversified but relatively focused companies tend to outperform highly diversified companies. And that finding has held up fairly consistently over the decades. Financial theory would say that companies diversify to reduce risk, but in the business world diversification is done not to hedge risk but to sustain top-line growth. The riskiest companies—the start-ups and early-stage companies—are intensely focused. Companies begin thinking about diversification only when their growth has plateaued and opportunities for expansion in the original business have been depleted. Suddenly, they have more cash flow than they know what to do with.

The Quarterly: Why are the highly diversified companies less profitable?

Richard Rumelt: It seems that the more complex an organization gets, the more likely it is that inefficient and unproductive businesses accumulate in the nooks and crannies and back alleys—and sometimes right up there in center aisle. These businesses are subsidized by their cousin, brother, and sister businesses that are doing well, and they stick around for too long because there’s a bias against shutting things down. Often we’ll find that these are pet projects of senior management and cutting them would cause a huge ego blow. It’s extremely unrewarding to a person’s career to weed the garden inside a company. It is much easier and more popular politically to grow the company than it is to go around and disrupt everybody’s neighborhood.

One of the things we see happening in private equity is highly incentivized people assuming this very unpleasant task of taking a company private, weeding its garden, and then taking it public again. It hasn’t happened with highly diversified companies yet, but we see that, essentially, something like that is happening as relatively complex organizations are cycling through private equity.

The Quarterly: In addition to not weeding the garden, are there other significant problems that you see senior executives failing to handle?

Richard Rumelt: Another one is the stock market. When I’m talking to CEOs, the subject that comes up over and over again is stock prices: how should CEOs deal with the pressures of the market?

There are two big problems with managing for stock prices. The first is that stock prices are extremely volatile—too volatile, really. Research has shown that lots of variation in stock prices has little to do with corporate or economy-wide performance. An engineer would say that the signal-to-noise ratio is very low.

The second big problem is that stock prices respond to changes in expectations, not to performance. When you improve your profits, the stock price does not necessarily move at all. It goes up only if the increased profit was a surprise. And if profits are up but not as much as expected, you can get dinged.

Now, that is simply not how most people think about performance. It violates our basic notions of fairness. If I treated my students that way, they would revolt! Suppose that at the beginning of the term I projected each student’s final-exam grade, taking into account IQs, grades in other courses, anything I could get my hands on. Then when the final rolled around, suppose that their students’ grades for the whole course reflected how well they did on the final relative to my initial expectations. Say that Alan scored 50 and Barbara scored 80 on the exam but that I gave Alan an A and Barbara a B because Alan did better than the 40 I expected and Barbara did worse than the projected 90. The students would riot, waving signs saying “unfair!” But that’s how the stock market works.

As a CEO, living with the stock market as a constant factor in your life takes iron nerve and an ability to be detached. You have to remind yourself that it is not measuring your recent performance; it is speculators adjusting their expectations about what will happen next.

The Quarterly: Instead of these distractions, what should CEOs be focusing on?

Richard Rumelt: The most important job of any manager is to break down a situation into challenges that subordinates can handle. In essence, the manager absorbs a good chunk of the ambiguity in the situation and gives much less ambiguous problems to others.

In a focused company, the CEO does this for the entire organization by examining the overall competitive situation and providing enough guidance to let the organization get to work. The CEO defines the business problem for everyone else.

In a diversified corporation, the CEO’s job is to keep the individual business units healthy. We know that the locus of success and failure is the business unit, not the corporation. The evidence shows that an average multibusiness corporation has little if any systematic effect on the businesses that it owns and manages. This is hard for many to hear, but it is a fact. So the senior management of the corporation should provide the resources and knowledge that each business needs to be healthy.

What makes a business unit healthy? Operating efficiently and having a good strategy. A good strategy, in turn, is one that is responsive to change and that builds, builds upon, and stretches the resources that yield competitive advantage.

The Quarterly: The resource-based view.

Richard Rumelt: Yes, the resource-based view, which at one level looks obvious. It says you’ve got to have good resources in order to have good results. But it’s really a theory about what’s the locus of success. Where is it coming from? It’s coming from having, within a company, difficult-to-replicate and usually intangible resources. Things that generate and sustain competitive success—things like reputation, a good customer group, network externalities, experienced and competent people performing your processes.

The Quarterly: How do you accomplish this in a world that’s changing so quickly? Does the very notion of a proprietary resource or a structural advantage have the same meaning that it used to?

Richard Rumelt: No, it doesn’t really. From the old learning-curve era—the experience curve era—we know that companies get good at something by doing it. It appears that by distributing and collecting DVDs, Netflix is getting good at doing that. Now, that doesn’t mean it was always good at it. We create our competencies by making bets and putting the right resources in place to develop those competencies. We have to understand that competencies are created by activity. If you internalize enough of those activities, you actually get good at them, and they give you a sustainable advantage for a certain period of time.

The Quarterly: Which is less than it used to be.

Richard Rumelt: Yes, and then the advantage evaporates on you.

Explore a career with us