Reorganizations: What to expect and avoid

Wharton experts examine recent megamergers and find that there are common themes to explain the success--and failure--of what are almost always jarring corporate restructurings.

10 min read
It's a familiar scenario: A company brings in a new department head who immediately decides that the way to show leadership is to reorganize. And then a new division head comes on board, or a new CEO, and there are more reorganizations, restructurings or reengineerings. Employees can find themselves reorganized several times in one month. No one, of course, would argue against the need for organizational restructuring, especially as companies become more global and expand into new products and markets. And few would argue against the long-held belief that corporate strategy and structure must be aligned in a way that allows the company to manage change efficiently.

The problems arise when reorganizations are undertaken for the wrong reason, are poorly implemented or fail to understand particular constraints of either the company or the market in which it operates.

"Deciding on the right strategy for the company is key," says John Paul MacDuffie, co-director of Wharton's Jones Center for Management Policy, Strategy and Organization. "Strategy-making processes done hastily and based on the wrong assumptions can mean moving the boxes around on the org. chart without thinking through all the consequences."

In a recent article entitled "Organizing to Deliver Solutions," Jay Galbraith, an organization design expert and former Wharton faculty member, talks about a trend in business strategy that involves offering solutions to customers instead of only stand-alone products--i.e., bundling products and adding software and services in ways that integrate customer needs and increase customer value.

"Companies always underestimate how difficult this process is, largely because it means creating a customer-centric view," Galbraith notes. "That is, if you are going to combine products and services for the benefit of the customer, you have to know a lot about who that customer is."

IBM, for example, integrates its growing service business with its software and hardware products into solutions for customers--such as a supply chain management information system or computer aided design for new product development, says Galbraith. Nokia is another example. The Finland-based cell phone manufacturer has created a customercentric front end unit--where specific customer strategies are formulated--and added it to the product-centric units at the back end of the company. Galbraith calls it a "front-back hybrid model."

His article discusses the processes required to link the "back-end product units to the customer-facing solutions units." But even well-managed companies, he notes, have trouble achieving this solutions approach. "It requires much more than bundling and cross-selling products." It requires major organizational change, including, for example, a new structure, new management processes, new measurement systems, new talent and new reward systems.

Sun Microsystems has been partially successful in creating solutions for portals and payment systems.

It also requires leadership that can interact with multiple product and customer unit managers as opposed to handling one business unit at a time. These are "tension-filled processes," Galbraith says, defining solutions strategy as both a "team sport and a conflict-resolving sport." Even IBM, which is a success story in this hybrid solutions approach, "still has to confront its mainframe people, software programmers and Web designers, all of whose lives revolve around their own separate products. Part of the issue is getting all of the products to work together."

Citigroup reorganized its commercial and investment banking part of the business during the mid-1980s through 1995, and "from 1995 on, it integrated with Travelers and made a successful transition to global industry groups and customers," says Galbraith. Less successful, he adds, has been Motorola, which added a global solutions unit that was not supported by the product units, and AOL Time Warner, whose efforts to integrate products and services didn't create enough added value for its customers. Sun Microsystems has been partially successful in creating solutions for portals and payment systems.

Galbraith and others note two recent trends in organizational restructuring. The first is the rise of dispersed headquarters, which Galbraith describes as "a tendency to put power and responsibility into whatever part of the world has the leading edge in a particular activity." Citigroup's foreign-exchange business, for example, is headquartered and managed out of London, the world's largest foreign-exchange market. Private banking is headquartered in Zurich, derivatives are in New York City.

"There are more and more countries around the world where you find advanced forms of knowledge, not just in your own country," says Galbraith. "Telecom is really driven by the Scandinavians and the Japanese, not by the U.S. Depending on what business you are in, there is always some place in the world that has to lead, and you need to be there."

A second trend is the move away from companies organized by geography in favor of organization by product line, with power shifting to those individuals responsible for global, rather than local, business units or functions. Also, with the rise of the global customer, account management requires greater cross-border coordination than was necessary in the past.

"Suppose I am a manager in a New York-based multinational bank in the early 1990s and I am trying to serve key global accounts like General Motors or General Electric," says Nathaniel Foote, a consultant at the Center for Organizational Fitness in Waltham, Mass. "But under the company's current structure, all the power resides in the country managers, who are probably asking themselves why they should serve GM when their local clients are much more important and the margins are higher."

However, as Foote and others acknowledge, the choice between organizing your company by geography or by product line (including the creation of global business units), is rarely clear-cut. An article two years ago in The Wall Street Journal Europe followed the path of Exide, the world's largest producer of automotive and industrial batteries, as it tried to reorganize in the face of heavy losses and increased competition.

According to the Journal article, Robert Lutz, CEO of Exide at the time, held five retreats between June 1999 and January 2000, in which he asked the company's top executives how they thought Exide should be organized. After much debate Lutz made the decision to "form six global business units primarily around its product lines. Most of its remaining country managers were demoted to the post of local coordinators...About half the company's top European managers resigned."

But, as the Journal reports, the new structure was short-lived, primarily because Exide bought an international battery maker and Lutz was afraid that one of the newly acquired company's top executives would leave if his operation became part of the new global business units. "Instead, Mr. Lutz tilted the structural balance back somewhat toward a geographic model" and the re-reorganization eventually became one of "blending the geography and product-line models."

Part of the trauma associated with reorganization stems from a change over the past decade in the traditional understanding between employers and employees.

In addition to Exide, companies including NCR, Ford Motor, and Procter & Gamble "have spent fortunes transforming themselves from" a structure based on geography to one based on product line, and most have ended up with a hybrid of the two," according to the Journal.

Opting for targeted, not sweeping change
Management experts also caution against focusing too much on structure rather than on processes, or as Foote puts it, "over-leading with structure" in the hope that it will suggest progress. "Structure--including reporting relationships, employee roles and responsibilities, authority over resources, etc.--is just one dimension of an organization," says Foote. "Within the existing structure, you can change people's incentives, offer retraining, adopt new information systems and work-flow processes, strengthen people's networks and so forth. It doesn't involve reorganizing but instead involves working to improve performance."

Foote uses the analogy of building a university, or any place that has multiple buildings. "You can either put all the paths down in advance, or you can have people walk and show you where the paths should be, and then lay the concrete."

Adds MacDuffie, "Processes can provide guidance as to what kind of structural change may be needed. The compelling concreteness of an org. chart can be deceiving, while a processes approach tends to be less rigid and may provide an opportunity for learning and adapting as you go along. At some point the need for a change in the formal organization might be revealed, but by then there are formal relationships in place to move that change forward."

MacDuffie cites the case of one multinational durable-goods company headquartered in Japan that was feeling pressure from managers to open a North American headquarters as well. The company resisted, MacDuffie says, opting instead to put a number of horizontal processes in place rather than a major reorganization. "That way the company could get procedures going that might at some point reveal the need for bigger changes, but in the meantime would allow new relationships to form and problem areas to clearly emerge. In other words, the company made some targeted changes rather than a big, sweeping structural one."

It's also important, when considering reorganizations, that companies "know their products and their markets," adds MacDuffie. "One of the things we concluded about the auto industry is that centralized R&D still makes a lot of sense. It's a very capital-intensive business and involves sophisticated technical expertise. It's not so easy to just have European, U.S. and Asian R&D facilities that are as capable as a centralized one. But if you are Nestle, or a similar consumer products company, and you have global brands but many local brands as well, then you probably want to have some functions on a decentralized basis, because that is the only way to really fit the local market."

Indeed, a recent article in the McKinsey Quarterly makes that point using GE as an example. "All too often, companies that reorganize merely copy the organizational charts of successful companies without recognizing that they may be operating under completely different conditions," the authors write. "The much admired flat organizational structure of GE, for example, is tailored to its strategy of operating only in mature and stable industries in which execution is the key to success. As a result, GE doesn't need a central R&D department or other central resources that often imply larger and more complex corporate structures."

But the article goes on to say that GE's model would be "quite wrong for a company pursuing a business idea in an innovation-driven industry such as high-technology or telecommunications, in which a powerful central R&D unit is essential to capture economies of scale and ensure that technology platforms are consistent across the company."

Another example is Hewlett-Packard, says Michael Useem, director of Wharton's Center for Leadership & Change Management. While the jury is still out on the long-term success of its merger with Compaq, Useem credits HP CEO Carly Fiorina with "extremely detailed premerger integration" during which the HP and Compaq teams spent "thousands of hours deciding what lines would be dropped, what positions would be changed and so forth. It's not that companies can't learn from each other's experiences," Useem says. "It's that smart leaders customize the reorganization to fit their own situation."

The same type of tailored approach to restructuring was evident in General Motors, IBM and American Express, all of which brought in new management in the early 1990s. "Each of those new management teams went through significant restructuring, and all have come out well," Useem says, "but if another company tried to replicate exactly what Jack Smith did at GM, or what Lou Gerstner and Harvey Golub did at IBM and American Express," they would have no such guarantee of success.

Part of the trauma associated with reorganization stems from a change over the past decade in the traditional understanding between employers and employees. According to Peter Cappelli, director of Wharton's Center for Human Resources, IBM in the 1980s "used to reorganize all the time. But the company also offered employment security, which meant that an employee might be asked to change locations or move to a different part of the organization, but he kept his job and his salary. Consequently employees tended not to resist these changes. Now, however, companies reorganize in ways that threaten people. Employees might not just be assigned to some other job, they might lose that job or be demoted. There are all kinds of negative consequences. It reflects a change in the way companies do business and the fact that they are not particularly inclined to protect employees."

When disruptions do occur, Foote adds, "it's like taking people who were playing soccer and suddenly telling them to play football. An employee has to learn a new position and all the rules and regulations that go with it."

Often reorganizations "are done in the name of taking out layers, and then perhaps putting in new layers," says MacDuffie. "There may be some efficiency-based gains, but you are also destroying some of the relationship networks that were already in place. Forcing the creation of new networks causes cynicism and risk aversion among employees who may feel it's not worth it to invest in another set of relationships only to have everything change once again."

Changing the culture
On Oct. 15, 1997, Charles Schwab CEO David Pottruck took his top 150 managers out to the Golden Gate Bridge, had them walk across it and then gave some speeches to rev up their enthusiasm for the company's wholesale move onto the Internet. "It was an attempt to engage people's emotions," says Useem.

Now, however, companies reorganize in ways that threaten people.

Reorganizations, Useem adds, are frequently difficult in part because of deep-seated cultural attitudes in the company that are hard to change. One way to make progress is to "reach into employees' hearts, as Pottruck did. Another is to take a thousand small tactical steps-- hiring people with a commitment to the new model or design, firing a few key people who are standing in the way, repeating over and over again why and how things are now different. After six months employees begin to realize that the company is serious about these changes."

IBM's Lou Gerstner has commented in speeches and in his book that the main point of restructuring the company was to establish a new culture, to change the way 300,000 employees viewed their jobs, says Useem. HP's Fiorina, he adds, said several months ago that part of the reason for announcing the Compaq acquisition was to facilitate a shake-up of the HP culture so that the kind of reorganization that was required could go forward, whether the merger went through or not.

"Organizations have to be dynamic, and people's roles are always going to shift," Foote adds. "The challenge is to be thoughtful about how and why you reorganize. Done well, it's a process of ongoing dynamic alignment."

To read more articles like this one, visit Knowledge@Wharton.

All materials copyright © 2003 of the Wharton School of the University of Pennsylvania.