Skip to Content

Stanford GSB News

 
  • Email
  • Print
  • Share

Economist Roberts Outlines Three Ways to Organize a Multinational

John Roberts, co-director of the School's Center for Global Business and the Economy is the author of The Modern Firm: Organizational Design for Performance and Growth, published by Oxford University Press in May 2004. Roberts delivered this lecture on a related topic in late 2003.

October 2003

STANFORD GRADUATE SCHOOL OF BUSINESS—In an increasingly global economy, multinational firms face a host of complex problems, and no one model can help firms resolve them. A company's specific history, environment, and strategy shape its organization and the management tools it uses to balance global trade-offs, economics Professor John Roberts told Stanford Business School alumni/ae Oct. 17 during an Alumni Weekend lecture sponsored by the School's Sloan Program.

One of the biggest challenges for multinational companies is finding a way to resolve continual trade-offs. Should they compete with low costs or quality? Should they focus on current performance or innovation for tomorrow? How can they balance individual initiative and cooperation? "Strategy is about choice," said Roberts. "It's as much about what you are going to do as what you aren't going to do. It involves a trade-off."

Firms respond to these challenges by organizing into one of three models: international (very centralized headquarters with a focus on efficiency), multi-local (units in different countries have a lot of autonomy with a focus on serving the local market), and global or transnational (a mix of multi-local and international).

Each model has its drawbacks. With centralized headquarters, for example, dealing with many or all of the world's 25 time zones makes managing from one place difficult. For multi-local companies, creating incentives and metrics for each individual branch is challenging. And global models require intensive communication and can be expensive to run.

So what model should a firm choose? That depends on the specific company, said Roberts, the John H. and Irene S. Scully Professor of Economics, Strategic Management, and International Business. If, for example, the firm's secret to success is its location, the company should stick with an international model and do as much as it can with exports and by tapping into the local competitive advantage.

If location is less important, the decision becomes trickier. Roberts pointed to two determining factors: Companies need to consider their transport and communication costs across regions, and whether preferences for the product are similar across boundaries. Preferences for a certain kind of computer chip, for instance, vary less than preferences for food.

Using these criteria, Roberts concluded that a domestic model works best in industries where it is very difficult to export and preferences are disparate. The international model should be used if the goods flow easily and preferences are similar. If there's some commonality in preferences but it's hard to move the goods around, multi-local models work best. In many cases, however, with falling trade barriers, easier communication, and cheaper transportation, the multi-local model is transforming into the global model.

Companies also use specific management tools to help resolve trade-offs. To find out exactly what works for successful firms, Stanford and McKinsey & Co. paired to study 16 multinational companies through in-depth interviews, data questionnaires, and managerial network studies. The study, dubbed the Globe Project and headed by Roberts, identified seven key "drivers" that companies use to help ensure the correct balance between global coordination and local responsiveness.

Companies fall into two camps, using either all "soft" or "hard" drivers. Soft drivers include more human touch management methods like strategy, networks, and culture, while hard drivers include process, metrics, and incentives. Firms across the board put an emphasis on structure, but to varying degrees.

Cement company Cemex, a massively centralized company, emerged as one of the archetypal hard companies, while 3M, known for its highly dispersed decision-making and innovation, came out as a classic soft company, Roberts said.

The study found that firms don't pull heavily on both types of drivers at the same time. For that reason, the future of 3M will be interesting to watch. The firm's new CEO, W. James McNerney Jr., came from General Electric, a company known to rely heavily on process, metrics, and incentives. He already has introduced Six Sigma, the quality improvement process used at GE. The implementation has boosted profits but could slow down innovation, a prospect that Roberts called "worrisome."

The most important factor to remember, however, is that each company requires unique consideration, Roberts said. "How you want to organize depends very much on the strategy you're trying to execute and the context in which you're trying to execute it."

—by Sarah Robertson

Related Reading

The Modern Firm: Organizational Design for Performance and Growth
John Roberts, Oxford University Press, May, 2004