Building Ambidexterity Into an Organization

The technological downturn, political turmoil and economic uncertainty of the last five years have reaffirmed to managers the importance of adaptability — the ability to move quickly toward new opportunities, to adjust to volatile markets and to avoid complacency. But while adaptability is important, it is not enough. Successful companies are not just nimble, innovative and proactive; they are also good at exploiting the value of their proprietary assets, rolling out existing business models quickly and taking the costs out of existing operations. They have, in other words, an equally important capability we call alignment — a clear sense of how value is being created in the short term and how activities should be coordinated and streamlined to deliver that value.

For a company to succeed over the long term, it needs to master both adaptability and alignment — an attribute that is sometimes referred to as ambidexterity.1 For example, Finland’s Nokia Corp. is trying out a vast array of new mobile technology offerings, while continuing to invest in its dominant handsets franchise. GlaxoSmithKline Plc is experimenting with alternative organization models, alliance partners and technologies in its search for new blockbuster drugs, and it is also pushing hard to maximize the return from its existing drug portfolio.

The trouble is, it’s difficult to find the right balance between adaptability and alignment. Focus too much on alignment and the short-term results will look good, but changes in the industry will blindside you sooner or later. Lloyds TSB Bank Plc, based in the United Kingdom, delivered spectacular shareholder returns throughout the 1980s and 1990s, in large part through CEO Brian Pitman’s single-minded focus on return on equity. But little attention was paid to understanding changing customer needs or to the morale of the workforce, which ultimately undermined the company’s performance. From 1998 to 2003, Lloyds TSB lost 60% of its market value.

Similarly, too much attention to the adaptability side of the equation means building tomorrow’s business at the expense of today’s. Consider the case of Sweden’s Ericsson, which has led the technological development of the mobile telephony industry. Ericsson developed one of the first analog mobile systems; it led the industrywide development of the global system for mobile communication; and it has pioneered general packet radio system and third-generation mobile technology standards. But the impressive growth in sales in Ericsson’s systems business masked a high-cost and bloated organizational structure. At its peak, the R&D organization employed 30,000 people in approximately 100 technology centers with considerable duplication of effort. Adaptability, in other words, had taken precedent over alignment, and the subsequent crash in the telecom industry meant that Ericsson was hit harder than most. Since its peak in 2000, Ericsson has laid off around 60,000 employees and closed most of its technology centers in a bid to restore the profitability of its current businesses.

Two Forms of Ambidexterity

The concept of organizational ambidexterity has been around for years, but the evidence suggests that many companies have struggled to apply it. The standard approach is to create structural ambidexterity, that is, to create separate structures for different types of activities.2 For example, the core business units are given responsibility for creating alignment with the existing products and markets; and the R&D department and business development group are given the job of prospecting for new markets, developing new technologies and keeping track of emerging industry trends. Structural separation is necessary, the argument goes, because the two sets of activities are so dramatically different that they cannot effectively coexist.

But separation also can lead to isolation, and many R&D and business-development groups have failed to get their ideas accepted because of their lack of linkages to the core businesses. Many companies have experimented with variants of the structural ambidexterity model. Some pull individuals out of their current jobs to work in a dedicated cross-functional team for a limited period of time. Others separate the different types of activities within a single business unit — for example, they create a small business-development team attached to a business unit. These approaches avoid the extreme form of separation that is typical of dual structures. But they nonetheless remain top-down in nature in that they rely on business-unit managers to judge how best to divide employees’ time between one set of activities and another.

In an attempt to shed new light on the phenomenon, we have developed and explored the concept of contextual ambidexterity,3 which calls for individual employees to make choices between alignment-oriented and adaptation-oriented activities in the context of their day-to-day work. (See “About the Research.”) In business units that are either solely aligned or solely adaptive, employees have clear mandates and are rewarded accordingly. But in a business unit that is ambidextrous, the systems and structures are more flexible, allowing employees to use their own judgment as to how they divide their time between adaptation-oriented and alignment-oriented activities. For example, should they focus on current customer accounts to meet quota, or should they nurture new customers with slightly different needs? To foster this sort of ambidexterity on the individual level, a much greater level of attention has to be paid to the human side of the organization.

Contextual ambidexterity differs from structural ambidexterity in many important ways (see “Structural Ambidexterity vs. Contextual Ambidexterity”), but the two approaches are best viewed as complementary. Indeed, many successful companies, including Hewlett-Packard, 3M and Intel, use a combination of both approaches to deliver simultaneously on the needs for alignment and adaptability.

Contextual Ambidexterity

Our research, which included interviews with a wide variety of employees, ranging from senior executives to front-line workers, identified four ambidextrous behaviors in individuals:

Ambidextrous individuals take the initiative and are alert to opportunities beyond the confines of their own jobs.

For example, a regional sales manager for a large computer company, in discussions with one large client, became aware of the need for a new software module that no company currently offered. Rather than try to sell the client something else or just pass the lead on to the business development team, he took it upon himself to work up a business case for the new module; once he received the go-ahead, he moved full time into the development of the product.

Ambidextrous individuals are cooperative and seek out opportunities to combine their efforts with others.

A large beverage company’s marketing manager for Italy was primarily involved in supporting a newly acquired subsidiary, and she was frustrated with the lack of contact she had with her peers in other countries. Rather than wait for someone at headquarters to act, she began discussions with peers in other countries that led to the creation of a European marketing forum. This group met quarterly to discuss issues, share best practices and collaborate on marketing plans.

Ambidextrous individuals are brokers, always looking to build internal linkages.

On a routine visit to the head office in St. Louis, a Canadian plant manager for a large consumer products company heard discussions about plans for a $10 million investment in a new tape manufacturing plant. He inquired further into these plans, and on his return to Canada, he called a regional manager in Manitoba, who he knew was looking for ways of building his business. With some generous support from the Manitoba government, the regional manager bid for, and ultimately won, the $10 million investment.

Ambidextrous individuals are multitaskers who are comfortable wearing more than one hat.

For example, the operations manager in France for a major coffee and tea distributor was initially charged with making that plant run as efficiently as possible, but he took it upon himself to identify new value-added services for his clients. He developed a dual role for himself, managing operations four days a week and on the fifth developing a promising electronic module that automatically reported impending problems inside a coffee vending machine. He arranged corporate funding, found a subcontractor to develop the software and then piloted the module in his own operations. The module worked so well that operations managers in several other countries subsequently adopted it.

These four attributes — which collectively describe an ambidextrous employee — have several important commonalities. First, they constitute acting outside the narrow confines of one’s job and taking actions in the broader interests of the organization. Second, they describe individuals who are sufficiently motivated and informed to act spontaneously, without seeking permission or support from their superiors. Third, they encourage action that involves adaptation to new opportunities but is clearly aligned with the overall strategy of the business. Such behaviors are the essence of ambidexterity — and they illustrate how a dual capacity for alignment and adaptability can be woven into the fabric of an organization on the individual level.

Still, an individual’s ability to exhibit ambidexterity is facilitated (or constrained) by the organizational context in which he or she operates, so contextual ambidexterity can also be diagnosed and understood as a higher-order organizational capability. At the organizational level, contextual ambidexterity can be defined as the collective orientation of the employees toward the simultaneous pursuit of alignment and adaptability. It is manifested in the behaviors of hundreds of individuals in the ways described above and in the unwritten routines that develop in organizations. In this respect it is analogous to the well-established concept of market orientation, which is a collective orientation of people throughout a business toward the gathering, interpretation and dissemination of market knowledge.4 And just as with market orientation, ambidexterity is a potentially important capability for contributing to long-term performance.

So what does an ambidextrous organization look like? There are numerous paths to ambidexterity, but consider two examples from companies whose units rated very high in our research on both contextual ambidexterity and performance:

Renault, the French automobile company, went through a radical transformation during the 1990s. When Louis Schweitzer became CEO in 1992, the state-owned company was languishing. Schweitzer cut costs through a number of well-publicized plant closures, but he also invested in new-product development (leading to such models as the Espace and Megane) and began the search for a strategic partner to take Renault into the top tier of the industry. After an abortive merger with Volvo in 1993, Renault gained control of a struggling Nissan in 1998 and, to the surprise of many observers, quickly turned around its performance. By 2001, the Renault-Nissan Alliance had joined the ranks of industry leaders and was one of the most profitable auto companies in the world.

How did the transformation take place? Schweitzer developed a simple and consistent strategy built around what he called the “seven strategic goals.” The strategic planning and budgeting processes, and the bonuses and stock option plans, were all aligned with these goals. The communication of the message was, in the words of one executive, “doggedly consistent.”

At the same time, the company developed what one executive called a “deep desire to adapt.” The seven strategic goals were updated every two or three years, the organization had an informal style of management in which expressing alternative views was encouraged and managers developed a self-critical approach, always looking to improve. The result was an organization that became proficient at continually making small adaptations to its strategy without losing alignment.

As a second example, Oracle Corp. is the leading enterprise software company in the world with more than $10 billion in revenues. Oracle’s rapid growth, and the continuing presence of its founder, Larry Ellison, created an entrepreneurial style of management that eschewed formal structures and processes wherever possible. And perhaps because of this, the concept of ambidexterity sits easily with its senior executives. As one of them put it, “We align around adaptability.”

The company has shown a “remarkable ability to turn on a dime.” Consider Oracle’s shift into e-business in 1999 and its current shift into services. Oracle achieved this adaptability by hiring very smart people, setting aggressive but not unrealistic targets and avoiding too much formalization. As one executive explained, “Moving at this high rate of speed makes it impossible to maintain formal processes. Instead, a lot of people are making unilateral decisions.” At the same time, though, the objectives, goal setting programs and incentive systems are carefully aligned. “Employees in all lines of business have a clear idea of the company’s objectives,” observed one manager.

Renault and Oracle are not alone. Tesco Plc, the leading grocery chain in the United Kingdom, delivers industry-leading profit margins through a well-aligned operational strategy while continuing to push the development of new store concepts and new product lines. 3M Co. is famous for its highly innovative work practices, also delivering impressive margins through its systematic financial control and continuous improvement systems. An ambidextrous organizational context can be achieved through a variety of means, but they all share one thing in common: They enable individuals in the organization to exhibit initiative, cooperation, brokering skills and multitasking abilities.

Building Contextual Ambidexterity

How can managers begin to think about building contextual ambidexterity into their organizations? Sumantra Ghoshal and Chris Bartlett define context as the often invisible set of stimuli and pressures that motivate people to act in a certain way.5 Along that line of thinking, top managers shape organizational context through the systems, incentives and controls they put in place, and through the actions they take on a day-to-day basis. It is then reinforced through the behaviors and attitudes of people throughout the organization.

Ghoshal and Bartlett argue that four sets of attributes —stretch, discipline, support and trust — interact to define an organization’s context. In combination, these attributes create two dimensions of organizational context: The first, performance management (a combination of stretch and discipline), is concerned with stimulating people to deliver high-quality results and making them accountable for their actions; the second, social support (a combination of support and trust), is concerned with providing people with the security and latitude they need to perform.

Performance management and social support are equally important and mutually reinforcing. The strong presence of each will create a high-performance organizational context that gives rise to a truly ambidextrous organization. However, if there is an imbalance in these organizational characteristics, or a lack of both, a less than optimal organizational context will exist. (See “Four Types of Organizational Context.”)

For example, a demanding, results-driven orientation that lacks social support will create a burnout context. Many people will perform well for a limited time in such a scenario, but its depersonalized, individualistic and authority-driven nature typically results in a high level of employee turnover, making ambidexterity difficult to achieve. Conversely, strong social support without high-performance expectations will engender a country-club context6 in which employees benefit from and enjoy a collegial environment but rarely produce up to their potential. Companies in this position also have low ambidexterity and produce satisfactory but lackluster results. An absence of both a high-performance ethic and social support will, of course, produce a low-performance organizational context. Employees are unlikely to be either aligned or adaptive, let alone ambidextrous.

Creating a High-Performance Organizational Context

While performance management and social support factors do not directly create high performance, they do shape the individual and collective behaviors that over time enable ambidexterity, which does lead to superior performance. The challenges of building such a high-performance context are illustrated by the cases of Renault and Oracle.

Renault’s transformation during the 1990s involved a shift from the country-club to the high-performance context. Until 1990, employees had viewed the company as a comfortable and secure place to work, with an informal atmosphere. Over the following 10 years, a number of changes were brought about, primarily through top-down initiatives revolving around cost reduction and quality and through greater focus on, and commitment to, key strategic objectives. One executive commented that his business unit was run as a “commando-type organization — appraisal and evaluation interviews are run in a pyramidal form and compensation is [now] geared toward short-term objectives.” Most of these changes were instituted through a new executive team that gave people more structure, which led to a focus on new products and new opportunities as a means of delivering on the more ambitious goals. Stated slightly differently, the emphasis during the transition was placed on performance management but building on the social support that had existed in the early 1990s. Indeed, two of Schweitzer’s seven goals were concerned with the internal organizational context (develop a coherent and open group; work more effectively together).

Compare this to Oracle, which was positioned on the cusp between the high-performance and burnout contexts. Performance expectations were very high, people were well rewarded and the style of working was competitive and aggressive. One executive, for example, compared the business to “the engine of a Ferrari, which revs at very high rpm but can burn out at any minute.” But at the same time, employees rated the company moderately high on social support, citing the development of a balanced-scorecard system and the leadership forums at which the top 275 managers gather to share ideas.

The contrast between these two companies raises three critical points:

First, there is no single pathway to ambidexterity: Renault achieved it by building a performance context around its existing social support; Oracle built a performance context first, then looked for ways of building support and trust across the organization.

Second, there is no single leadership model for an ambidextrous organization. Larry Ellison is charismatic and directive; Louis Schweitzer is no less powerful, but he works in a more collegial manner.

Third, despite all their differences, Renault and Oracle both exhibit a clear and simple set of priorities. In our survey analysis, Oracle employees emphasized goal setting, individual performance appraisal and risk management; Renault employees highlighted capital allocation, recruiting and vision. Obviously, the choice of focal elements is important, but even more important, our evidence suggests, is the consistency with which they are applied and the number of employees they affect.

Escaping From Suboptimal Contexts

Many companies find themselves mired in contexts that do not effectively support ambidexterity and high performance. Those companies have to look for ways to engineer dramatic shifts in the behaviors they encourage.

The burnout context, for example, puts so much emphasis on performance management that social support systems are either neglected or never put in place. Eventually, performance suffers as exhausted and disenchanted employees have neither the capacity nor the incentive to execute or innovate. Clifford Chance Llp, the world’s largest law firm, had clearly reached this stage in October 2002 when an internal memo was leaked to the press revealing that staff in the New York office were expected to bill 2,420 hours per year —or roughly 10 hours per day. The memo explicitly stated that “the stress of billable hours is dehumanizing and verging on an abdication of our professional responsibilities.”

Scotch Inc. (not its real name), one of the largest consumer products companies in the world, provides another example of burnout. Scotch had grown quickly during the late 1990s through a strategy of focusing on a small number of core brands and rolling them out quickly on a global basis. By 2000, however, growth was slowing, and the foreign subsidiary managers were starting to voice some concerns. The managers had limited influence over the positioning of the global brands in their local markets; they were short of resources; and they felt the strategic planning process was too top-down. At the same time, the growth goals were demanding, and there was little or no tolerance for failure. The emphasis on performance management had led to solid growth, but executives were concerned about where the next phase of growth would come from. Scotch senior management recognized the potential for burnout and introduced several initiatives to increase the quality social support offered by the organizational context, including changes to the strategic planning process, introduction of systems for sharing best practices among subsidiaries and refinement of programs for professional development.

The country-club context — in which there is a strong sense of support and trust, but no one works too hard and mediocre performance is tolerated — can be as dysfunctional as the burnout context. Many government agencies, universities and state-owned companies fall naturally into this category as do a fair number of commercial organizations. For example, Lufthansa AG had such a culture during the late 1980s in large part because, as the state-owned airline, it was too important to fail. When Jurgen Weber took over as CEO in 1991, however, he clearly demonstrated that the company was very close to bankruptcy. He put a performance management dimension in place, and the context began to shift toward high performance.7

Similarly, Cowes Ltd. (not its real name), a formerly state-owned European dairy-products company, sold farm produce to consumer goods companies, typically within a strict set of regulations and quotas. Faced with impending deregulation, however, Cowes’ country-club culture put the company in a noncompetitive position. Senior executives realized this and made two significant changes: They broke the company down into distinct profit and loss units and instituted a pay-for-performance scheme for unit managers, and they introduced innovation processes to seek out new sources of top-line growth.

Although the strategies employed by Scotch, Lufthansa and Cowes were initiated in a top-down fashion, each sought — by emphasizing performance management alongside social support — to create a high-performance context in which ambidextrous behavior on the individual level would be encouraged and rewarded.

At some companies, there is not only little concern for performance, but also no sense of trust or support among the employees. That is the worst of both the country-club and burnout syndromes and constitutes a low-performance context in which ambidexterity is impossible. For example in one company we studied, there was evidence of inconsistent messages from top management (making it hard to create trust) and a sense that the business lacked the ambition or focus needed to generate stretch. As one manager said, “There is no overarching vision, each division devises its own vision and objectives.” In another company, there was evidence of a lack of follow-through when using management systems (making it hard to create discipline). For example, there had been a number of new initiatives, which according to one manager “had lost accountability and steam” within less than a year. Support systems, in terms of providing training, feedback sessions and information across the functions, could be identified in both companies, but were insufficient on their own to develop an effective organizational context.

Companies that find themselves in a low-performance context must place an immediate priority on developing improved performance management. Social support mechanisms can follow well before the risk of burnout becomes an issue.

Pathways to Ambidexterity

For executives who are seeking to build an ambidextrous organization, there are five key lessons that emerge from our work.

Diagnose your organizational context.

Before an organization can take steps toward a high-performance context, it must discover where it currently stands in terms of performance management, social support and the balance between the two. A simple diagnostic tool (see “Diagnosing Your Organizational Context.”) that involves responses from a large cross-section of people throughout the company will produce a basic, helpful quantitative analysis. It can be supplemented with a more qualitative discussion of the context in the organization. To the extent that the two analyses reinforce each other, a reliable picture emerges of what changes need to be made along what lines to move the organization toward high performance.

Focus on a few levers, and employ them consistently.

We found no evidence that specific organizational levers, such as incentive compensation or risk management, were consistently linked to success. There are many ways to build an organizational context that enables ambidexterity. The higher-performing companies, however, are those that focus consistently on just a few levers. For example, Scotch decided to focus on levers intended to bolster its social support side of the equation: professional development, knowledge transfer and a more participative strategic-planning process. The more consistently those are applied, the easier it will be for employees across the organization to make sense of the changes that are under way. Consistency is crucial since organizational context does not, on its own, create high performance but enables the individual-level ambidexterity that, over time, leads to high performance.

Build understanding at all levels of the company.

In our survey research, we found that the lower the respondent was in the corporate hierarchy, the lower he or she rated the organization’s ambidextrous characteristics — a pattern we call the erosion effect. Intriguingly, the magnitude of the erosion effect varied with the performance of the company. In lower-performing companies, front-line employees rated elements of organizational context an average of 1.5 points lower on a 7-point scale than did their top-management counterparts. In the more ambidextrous and higher-performing companies, the rating disparity was typically 0.5 points or less.

The interview component of the study revealed that the erosion effect is a measure of the consistency and quality of communication in the organization. That is, for organizational context to be effective in creating ambidexterity, its message has to be disseminated clearly and consistently throughout the organization. Unless lower-level employees genuinely understand the initiatives of top management, the initiatives will have a minimal impact on individuals’ capacity for ambidexterity.

View contextual ambidexterity and structural ambidexterity as complements.

Almost all the previous research on ambidexterity has focused on structural separation between alignment-oriented and adaptability-oriented activities. There is evidence that this approach can be highly effective, but there is also evidence that it can create as many problems as it solves. For example, many large companies, including British American Tobacco, Royal & SunAlliance and British Airways, established corporate venture units during the dot-com boom to nurture new business ideas, but as those units lacked sufficient connective tissue with the core business, most eventually became isolated and irrelevant to the company’s strategy.8

Contextual ambidexterity isn’t an alternative to structural ambidexterity but rather a complement. Structural separation may at times be essential, but it should also be temporary, a means to give a new initiative the space and resources to get started. The eventual goal should be reintegration with the mainstream organization as quickly as possible. Contextual ambidexterity can enhance both the separation and reintegration processes.

View contextual ambidexterity initiatives as “driving leadership,” not as being “leadership-driven.”

Ambidexterity arises not just through formal structure or through the vision statements of a charismatic leader. Rather, it is achieved in large part through the creation of a supportive context in which individuals make their own choices about how and where to focus their energies. Leadership, in other words, becomes a characteristic displayed by everyone in the organization.9 The impetus toward ambidexterity may sometimes be driven by top-down initiatives, but the goal is to allow leadership to emerge from the organization at all levels and for that ubiquitous, emergent leadership to be inherently ambidextrous.