All companies hit rough patches from time to time. But only a few manage to survive decade after decade — some of them in a form that bears no resemblance to the original organization. Nokia began in 1865 as a riverside paper mill along the Tammerkoski Rapids in southwestern Finland. In the late 1880s, Johnson & Johnson got its start by manufacturing the first commercial sterile surgical dressings and first-aid kits. And in 1924, the founder of Toyota came out with his company's first invention — an automatic loom.
What explains this longevity? Stanford Graduate School of Business Professor Charles O'Reilly calls it "organizational ambidexterity": the ability of a company to manage its current business while simultaneously preparing for changing conditions. "You often see successful organizations failing, and it's not obvious why they should fail," O'Reilly says. The reason, he says, is that a strategy that had been successful within the context of a particular time and place may suddenly be all wrong once the world changes.
Staying competitive, then, means changing what you're doing. But the change can't be an abrupt switch from old to new — from print to digital distribution, say, or from selling products to selling services — if that means abandoning a business that's still profitable. Hence the call for ambidexterity. You can't just choose between exploiting your current opportunities and exploring new ones; you have to do both. And the companies that last for decades are able to do so time and time again.
O'Reilly's work builds on that of other organizational scholars who have noted the value of a two-pronged survival strategy. In a seminal paper published in 1991, Stanford Professor James March wrote about the need for organizations to do two things at once, and articulated the challenge. "Both exploration and exploitation are essential for organizations," March wrote, "but they compete for scarce resources." That means organizations that try to do both face difficult trade-offs, choosing one only at the expense of the other. Harvard Professor Clayton Christensen went a step further, pointing out in The Innovator's Dilemma in 2011 that the very things that make an organization successful today will actually work against it as conditions change. It's not just that resting on your laurels is tempting, or that managers are blind to the changes around them. Rather, innovation can easily seem like a threat to a business that is already working well.
When Christensen wrote The Innovator's Dilemma, he saw no way out, O'Reilly says, except to spin out the innovative part of the organization. According to that approach, the best way for Wal-Mart Stores Inc., for example, to cope with the advent of internet retailing was to continue to focus on its brick-and-mortar stores and to spin off website Walmart.com as a separate company, as it did in 2000.
But a spinoff doesn't really solve the problem, O'Reilly says, because it doesn't help Wal-Mart make money in the long run. A better way, his research suggests, is to run the mature business alongside the newer business under the same organization — but, crucially, to do it in a way that makes smart use of the organization's resources.
A good model is the way in which Wal-Mart is rolling out its Express stores, the much smaller alternatives to the company's behemoth supercenters and among its best hopes for continued growth. This venture, which is moving in on the turf occupied by the likes of CVS and Walgreen, seems likely to pay off, O'Reilly says, because Wal-Mart's senior managers aren't merely moving into a new, related business; they're leveraging "the strengths of the mother ship" to do so. For Wal-Mart, those strengths are in real estate, purchasing, logistics, and information technology — all capabilities that will be useful in the drugstore business, too.
Christensen, O'Reilly says, now sees ambidexterity as the solution to the innovator's dilemma, but not everybody does. The idea that organizations can reshape themselves to adapt to change runs counter to a decades-old tradition in organizational studies that says, in effect, that organizational survival is a matter of luck. That school of thought, influenced by evolutionary theory and known as organizational ecology, holds that the companies that survive today are products of natural selection. These organizations have the right features to thrive in their current environment, organizational ecologists say, but sooner or later, the environment is bound to change. And if it changes in ways that favor a different set of traits, the argument goes, an individual business can't adapt any more than a zebra can change its stripes.
That view is too fatalistic, O'Reilly believes, because it ignores managers' power to learn and change. If Wal-Mart is continuing to grow while Sears is in decline, it's because Wal-Mart's leaders are deliberately doing the right things.
O'Reilly and his colleagues, especially his close collaborator Michael Tushman, of Harvard Business School, have found what some of those things are. Above all, an ambidextrous organization needs a leader with an "overarching vision," or clarity about why different businesses within the organization are important. But their research also shows that problems arise when other senior managers disagree with that vision. Therefore, the leader must also "make sure that everybody is singing off the same hymnal," O'Reilly says.
Managers must make sure their organizations actually align with that vision, as well — a difficult feat, given that different business units' cultures and incentives might be tugging them in different directions.
The best leaders manage to pull it off. One example is Glen Bradley, who in the early 1990s led Ciba Vision, a maker of contact lenses that was losing ground to Johnson & Johnson. Johnson & Johnson had the economies of scale to defeat Ciba Vision in the market for conventional lenses, so Bradley redirected his organization's resources toward developing innovations, such as contacts that people could wear while sleeping. At the time, the concept of extended-wear contact lenses was to conventional contacts what digital photography had been to Kodak's film business: If successful, many feared, the new product would kill the old one.
To make clear why the old business should support the exploratory projects, Bradley crafted a new vision for the entire company: "Healthy Eyes for Life," a statement whose breadth conveys the idea that the company should pursue whatever technologies and opportunities they had to promote healthy eyes. To forestall conflicts over resources, he set up a separate organization for each project, each with its own research and development, marketing, and finance group, and each headed by a leader given free rein to create the right culture to meet that organization's goals.
At the same time, Bradley wanted to make sure the new projects benefited from the expertise of the old business, so he put all of them under the control of a single executive, who knew the old business and had the personal relationships to facilitate sharing across divisional boundaries. Bradley also revamped the company's incentive systems, to reward managers mainly for the performance of Ciba Vision as a whole. Thanks to these efforts, the new project teams became remarkably productive: Besides new types of contact lenses, Ciba Vision successfully introduced a drug to fight eye disease and pioneered a manufacturing process that greatly reduced the cost of making lenses. In the first 10 years after Bradley's move to ambidexterity, the company's annual revenues grew from $300 million to more than $1 billion.
Ciba's experience shows that with deft ambidextrous leadership, an underdog can stand up to a powerful rival. But Johnson & Johnson could have done what Ciba did. We often think of large organizations as lumbering bureaucracies incapable of swift change, a notion perpetuated by highly visible David-and-Goliath stories in business. (Think Netflix trouncing Blockbuster, which had years to respond to the little company with the red mailers.) In fact, large companies are often better-positioned for ambidexterity than small ones, O'Reilly says, because one bad bet won't wipe them out.
"If you're a small company, you place all your chips on this one thing, whereas a large organization can do lots of experiments," he explains.
IBM, an organization that O'Reilly has studied extensively (and for which he and Tushman have consulted), is a case in point. In 2000, the company's leaders, acknowledging that running their existing businesses with incremental improvements wasn't enough to grow revenue, launched a project to foster more exploration. Called Emerging Business Opportunities, the initiative might sound like just another stuffy big-company acronym. But reading O'Reilly's descriptions of the EBOs makes them look almost like startups within Big Blue, with each reporting to a division head and to the head of new growth opportunities — somewhat the way entrepreneurs remain accountable to their funders. Like actual startups, some of these organizations failed to bear fruit. But there were enough of them (seven in the beginning) that in the first five years alone, the EBOs added $15.2 billion to IBM's top line, O'Reilly and his colleagues report, or more than twice as much as acquisitions did.
A recent study by O'Reilly and colleagues suggests that while IBM's experience was extraordinary, the company does have something in common with other thriving organizations. The researchers looked specifically at what type of corporate culture was associated with growth in revenue and net income, and found that more adaptive cultures, or ones that emphasized speed and experimentation, did much better. "A culture that says, 'We don't have all the answers; we've got to try these experiments' — that's the type of culture that promotes ambidexterity."
What determines the ideal balance between exploration and exploitation is one of the big open questions in the research on ambidextrous organizations. It's safe to say, though, that the right amount of experimentation has much to do not only with a company's resources, but also with the pace of change in its industry. "If the industry isn't changing rapidly, doing 100 experiments is unproductive and expensive. But if you don't do experiments, you're likely to be in trouble if the industry is changing."
Charles O'Reilly is Stanford GSB's Frank E. Buck Professor of Management.
James March is the Jack Steele Parker Professor of International Management, emeritus.