The Limits of Structural Change

“At the end of the day, the essence of a company is not what you do — it’s what you know.”1 GARY HAMEL

In the last decade of the 20th century, businesses devoted considerable energy determining and structuring “what they do.” In addressing business problems, the tool of choice for managers has often been periodic reorganization — to enable them to better serve customers, product markets and channel partners. They’ve created structures that simplify execution of key processes and have even invested effort to optimize and reoptimize organization designs, right down to the project level.

Indeed, generations of leaders have looked first to changing organizational structure as a way to improve business performance. In the 1920s, Alfred P. Sloan and others articulated a doctrine that came to be known as the “3 S’s” — having crafted a strategy, senior managers must find a structure to fit it and align the two with supporting systems — and this logic became widely taught in the American academy.

In light of the speed of change in today’s economy, however, this view has come in for some criticism, perhaps epitomized by Michael Hammer who, with unintended irony, disparages “the widespread malady of ‘structuritis,’ whose principal symptom is the propensity to issue a new organization chart as the first solution to any business problem.”2

The pattern of frequent restructuring, prevalent across industries, is not without logic. Over the years, cement and paper manufacturers have migrated from functional to regional profit-center organizations to better manage their difficult constraints of low-value commodities and high transport costs. Conversely, semiconductor companies have evolved into product organizations, as their products’ more favorable economics allow them to serve multiple geographic regions without maintaining all the costs of a local presence in each one. Elevator service companies, which used to organize regionally to help supply local needs for service and spare parts, no longer need to do so, now that cab failures are both less frequent and more manageable through the use of remote electronic sensors.3

On the one hand, all of this restructuring has had substantial ameliorative effects on business performance. On the cost side, suitably aligned structures have often achieved the cost efficiencies they’ve targeted. On the revenue side, service improvements have generated more satisfied customers and commensurately increased orders. Internally, many employees’ exposure to diverse areas of the business within these fluid structures has deepened their knowledge base, to the benefit of both themselves and their organizations.

On the other hand, the use of organizational structure as a primary lever to solve business problems has coincided with, and been partially responsible for, a time of profound disillusion, particularly with large corporations. Matrix organizations have been widely derided. Virtual organizations have been successful at creating flexibility, but the people working in them struggle with loss of proprietary knowledge and control. And professional employees — perhaps reshuffled one too many times in the constant churn, or ushered out of companies entirely due to short-term exigencies — are showing demonstrable loss of loyalty and commitment to their employers, often rethinking their professional identities entirely.

In short, the execution of structural change has been suboptimal. For a variety of reasons rooted in the dynamics of business systems and in basic human nature, companies have missed opportunities to accomplish more in their restructuring initiatives.

However, as regular restructurings proceed apace, something more profound may be happening. Corporations, in their rush through numerous generations of “all the rage” organization designs — from functional “command and control” to flat, matrix and virtual structures — may now have perhaps inadvertently reached the final stage of organizational evolution. Against the backdrop of revolutions in technology, learning and career development, some leaders, recognizing the inherent limitations of structural change, are consigning organizational redesign to the “back burner,” and are choosing instead to spark change with more long-term, potent catalysts.

The Declining Prominence of Hierarchy

The frequency with which organizations restructure themselves implicitly acknowledges that restructuring benefits have been hard to achieve or fleeting when achieved. In that light, it seems not too radical to ask what purposes hierarchies — whether vertical or flat — served in the first place, and to what extent they continue to serve those purposes today.

The organization chart originally evolved as a response to the need for organizational order. It served as a framework to enable or optimize a variety of functions, including the control and separation of duties, top-down resource allocation, ownership of knowledge, the cascading of communications and the management of performance, compensation and careers.

Current technology capabilities and today’s increasingly individualistic approach to career planning, however, have rendered many of these functions outdated, or else provided alternative methods of managing them — ones that depend less on hierarchy (see “Beyond Hierarchy.”).

For example, organizational communication was traditionally controlled and colored as it cascaded from level to level in the hierarchy. Now, however, technology such as e-mail, voice mail, instant messaging and the Internet has disintermediated communications, and in so doing helps prevent the miscommunications associated with the handoff of messages from person to person across levels.

Performance management is another function for which many companies no longer have to depend as heavily on the reporting hierarchy. As knowledge workers have become the predominant employee group and cross-departmental project work has become prevalent, performance management is increasingly a matter of project-by-project evaluations, often done by project managers who are not those employees’ bosses. Employees themselves often consolidate their individual project evaluations, noting common feedback trends, and only involve their sponsors at the end of the process as to what the feedback should imply for their reviews, rewards and development plans.

How Compulsive Restructuring Affects People

Indeed, it’s apparent that hierarchy is playing a significantly smaller role in the “nuts and bolts” of how work gets done, and many of the original rationales for giving it primacy have become weakened. In that vein, it is a bit ironic that thousands of companies continue to spend millions of dollars continuously trying to identify an optimal hierarchy — one that employees then frequently proceed to ignore in favor of operating within their informal networks.

Take, for example, a restructuring in which responsibility for companywide IT initiatives, which had been previously housed in multiple strategic business units, is moved into a shared services center. The structural change is logical, as it reduces duplication, achieves scale economies and provides a more attractive work environment for professional IT managers. However, no restructuring is perfect, and the new department also has the potential to move critical decision making away from the work, become an internal monopoly, degrade intraorganizational responsiveness, and, in so doing, create perverse incentives for departments to “work around” the shared services center.

If such problems were to arise, a typical corporate response might be yet another reorganization (indeed, many managers focus on installing new structures rather than on building effective processes). A better solution, however, would involve keeping the shared services structure and the efficiencies it offers, while creating more of a results orientation by focusing on nonstructural issues such as people, processes and rewards. For instance, managers could be rotated between the central unit and strategic business units on a regular basis to diversify their experience and foster mutual appreciation and intracompany efficiency. In addition, corporate heads could reinforce service-level agreements in order to hold the central unit to rigorous responsiveness goals and could offer a more meaningful portion of bonus pay to unit members for meeting those goals.4

When organizations facing adversity and the complex process of managing human change habitually revert to rejuggling hierarchy and laying off key people, a troublesome dynamic arises —especially when those companies ignore other obvious palliatives. Widespread layoffs and confusion about roles, processes, compensation systems and the ground rules for work that are associated with the decreasing relevance of hierarchy all contribute to a fundamental breakdown in employee loyalty and a pervasive alienation and cynicism.

The obvious paradox here is that we ostensibly live in an era of employee “ownership.” Today’s shareholders are more aptly described as “investors” (or even “gamblers”) than “owners,” and most people wouldn’t dispute that the real value of companies like Intel Corp. or Pfizer Inc. lies less in their capital assets than in the heads of their employees. Yet shareholders rather than employees seem to hold all the cards in today’s rapidly restructuring economy.5 Consider the recent comment to analysts by the chairman of the Goldman Sachs Group (for which he later apologized to employees by voice mail): “I don’t want to sound heartless,” he said, “but in almost every one of our businesses, there are 15 or 20% of the people that really add 80% of the value, and if market conditions worsen, we will take more people out.”6 And for every public snafu like that one, there are likely many internal ones.

Arising out of employees’ restructuring-driven confusion and cynicism, we are now seeing the growth of new forms of association somewhat akin to the growth in unionism during the Industrial Age. Middle managers and even executives are becoming ardent networkers, alumni group participants and professional association joiners. These thriving networking groups provide a nonthreatening environment — often virtual — for communication, idea sharing and interaction, in ways that can be very beneficial. In so doing, they are filling roles once filled by employers: providing identity, community and support for their members. It may no longer be possible for managers to assume that people are a fungible, bottomless pit of available resources to be used and disposed of as a first reaction to market events. Especially as the economy improves, companies will be constrained by the needs of an increasingly mobile, empowered and educated workforce.

Alternative Catalysts of Organizational Change

This changing dynamic between employer and employee, coupled with the declining prominence of hierarchy, is necessarily shifting the emphasis away from restructuring as the first tool employed to effect organizational change.

As a cautionary tale, consider the attempt by one large engineering firm to facilitate organizationwide learning and development. When, in 1998, the company sought to more closely integrate its work in practice areas including urban design, survey and land development across its offices in 30 U.S. cities, it implemented a classic matrix structure. In this reorganization, executives at the helm of each of the vertical practice areas assumed nationwide responsibility and served as peers to city leaders working across practices.

However, few new compensation programs or process redesigns accompanied the change. As a result, according to a region head, implementation issues still abound, five years into the process. Superb geographical performance in some local markets, unfortunately accompanied by below-average corporate performance, has left many high-performers without bonus recognition. The performance management process, unrevised to reflect new reporting relationships, has not effectively communicated to professionals the importance of their role in the overall enterprise. Significant tensions prevail across the business line/geography axes regarding the hiring, cross-training and career development of employees. And many managers don’t have the means to identify the right resources for the right projects, or, if they do, to reward them effectively.

The redesigned matrix was a typically sensible “anatomical” restructuring, but was inadequately supported by the “physiology” of process, rewards and people initiatives to help address inevitable structural conflicts. This was especially problematic in an organization that sells its talent and skill sets.

In contrast, many companies have come to realize that periodically recasting a hierarchical mold — one that will inevitably be made obsolete by the fast-changing environment — is a losing battle. These companies have simply shifted their change management focus from the anatomical to the physiological.

Knowledge Management Disintegrates Walls at BP

Traditional company hierarchies derived a certain power and control from the limited dispersion of knowledge throughout the organization. Department managers tackled projects using the resources under their hierarchical control, often disregarding more appropriate resources with more specialized knowledge that might exist elsewhere in the company. Similarly, if an employee developed a certain expertise that would be of particular value outside his or her work group, that information would not be widely known. As a result, the organization as a whole would derive limited utility from that expertise.

Today, many companies (large service firms, in particular) have come to realize that if they can’t marshal broad-based resources to solve complex problems — a capability referred to by Morten T. Hansen and Bolko von Oetinger as “epiphanies of scale”7 — then they have no reason to exist as large entities. As a result, many have launched formal and informal knowledge management applications and processes so as to break through the constraints of hierarchy, disperse awareness more broadly throughout the organization and make the right talent and the right resources available for any given initiative.

BP Plc (formerly known as British Petroleum) has been so successful in this regard that its managers are now given incentives to not only manage vertically but also to provide horizontal “peer assists” to others throughout the global organization.8 A generation more advanced than the classic matrix, BP’s organization effectively mobilizes the right resources for the right job, regardless of their location or hierarchy. When a series of lightning strikes damaged pumping facilities in Siberia, local engineers, inexperienced in dealing with this type of problem and facing a bewildering array of remedial options, were quickly routed to valuable guidance from a Houston expert and participants in a Canadian network forum on the subject. Countless other tales illustrate the mobilization of global talent to increase the efficiency of retail operations, evaluate tanker-financing options and launch new plants in remote areas. Connectivity and assistance are readily available in the company’s marketplace of experience, and, within reason, it’s expected that cross-unit assistance be asked for and provided.

“BP’s approach represents a subtle but vital shift in the sources of advantage for large global companies,” write Hansen and von Oetinger. “In the past, the key advantage large companies had was their ability to pool volume across business units and countries to lower purchasing, component and production costs and better leverage their brand. These economies are still important, but the benefits of cross-unit learning and collaboration have become much more important.”9

Performance Measurement Obviates Hierarchy at Duke Power

Modern performance measurement systems and technologies —with their ability to house untold riches of transaction-level information, sliced and diced by any parameter imaginable —have played no small part in the rise of matrix organizational structures. Product managers now have relative product profitability information at their fingertips, customer segment managers are intimately familiar with relative customer profitability figures and so on. However, while such sophisticated measurement systems have made possible a granular management capability unheard of a decade ago, they also have created conflict. Should we discount a key product line as a loss leader to build this market? Yes, says the market head responsible for overall market growth. No, says the product head responsible for his product’s margins.

Duke Power Co., an electric power provider based in Charlotte, North Carolina, has managed to create a system to reconcile such structural conflicts — and through this has largely relegated the importance of hierarchy itself. Its strategy for performance improvement has been to cross-align performance measures and targets in such a way that market heads have product margin responsibility, and product heads are partially evaluated on market performance. In this approach, process owners and regional heads are positioned as peers, but have to share responsibility for the effectiveness of each process and the performance of each market. Common measures and targets force cooperation among diverse managers.

As an example, in other electric power organizations with narrower performance measures, process managers for line installation and maintenance might be at cross-purposes as each tries to optimize his own processes. As those processes are usually executed by the same people, the managers must compete for resources. But at Duke Power, broadly based market-level incentives induce creative resource allocation solutions, whereby installation and maintenance jobs are flexibly assigned to employees based on seasonality of work volume. This sometimes works to the temporary detriment of one process or another, but is always to overall benefit.

Moreover, Duke Power’s measurement systems actually render traditional notions of anatomical hierarchy relatively meaningless, as they blur and make nearly irrelevant the question of to whom the employees report. Each employee gets more or less the same guidance whether they go to a process head or regional manager for assistance. So the question “For whom do you work?” is often answered with, “It doesn’t matter.”10

Antistructure Cultivates Responsibility at W.L. Gore

Some companies don’t even use organization charts in the traditional sense of the term. W.L. Gore and Associates Inc., the maker of Gore-Tex fabric, is organized as a loose network of employees responsible to each other and their projects but lacking job titles. Bill Gore, the company founder, dubs this a “lattice” structure.11 Natural leaders obviously do emerge in the system, but that happens organically rather than by appointment. It is the job of what the company calls “advocate sponsors” to make sure that associates are appropriately recognized for their accomplishments and contributions to the company; responsibility and knowledge are replacing formal hierarchy in their decision-making processes.

Other companies have adopted similar approaches. Employees of several management consulting firms, for instance, although they retain the use of job titles, effectively have organizational boxes but no formal reporting lines. Sibson Consulting and Booz Allen Hamilton are two examples. Select senior staff act as “coaches” who support performance management and compensation by distilling associates’ various project reviews, and working with each associate to consolidate organizational feedback and attach appropriate rewards.

The End of Formal Structure

To ask how an organization should be formally structured is thus to ask an outdated question. Christopher Bartlett and Sumantra Ghoshal noted years ago that, “as many companies have discovered, reconfiguring the formal structure is a blunt and sometimes brutal instrument of change. A new structure creates new and presumably more useful managerial ties, but those can take months and often years to evolve into effective knowledge-generating and decision-making relationships.”12 The trouble is that in today’s economy, that is just too long.

While structure is clearly still part of the equation, strategists are now being forced to change their image of “organization” from that of a static photograph of boxes and lines to a dynamic movie depicting a continuous flow of internal and external responsiveness. Flexibility is trumping structure as the governing principle behind organization design. Physiology is trumping anatomy.

The culture change engendered by effective knowledge and performance management systems, and other people, process and rewards initiatives will reverse the degradation in trust that professional employees have experienced during the era of structural pre-eminence and periodic reorganization. Freed from the pigeonholing of their reporting lines (that never reflected the reality of their working relationships anyway), this change will energize professionals to more freely work cross-functionally and with ever evolving groups of new colleagues. Since competitive advantage will depend not on a formal structural order, but on the agility of the organization and the adaptability of its processes, people and technology, companies will truly be able to use “what they know” to build lasting value.