12 Jan 2004  Research & Ideas

Does Your HQ Operation Fit With Corporate Strategy?

Is a lean headquarters operation the key to success? How should headquarters design fit with corporate strategy? New research from professor David J. Collis has surprising answers.


The specter of "headquarters" usually looms large in the business world's imagination, but few managers—and few CEOs, especially new CEOs—understand how size, structure, and performance of headquarters interact. To investigate, Harvard Business School senior lecturer David J. Collis joined forces with colleagues David Young and Michael Goold on a wide-reaching project. With five goals in mind, which Collis explains below in an e-mail interview with HBS Working Knowledge, they surveyed 600 corporations in seven countries, and published the results in a working paper, "The Size, Structure and Performance of Corporate Headquarters." Among their findings: Contrary to popular belief, European HQs were smaller than American HQs. And the much-touted "lean and mean" ideal does not necessarily lead to better financial performance.

Martha Lagace: What was it about headquarters that drove you to study them further?

David J. Collis: Little academic research has been done on the topic even though it is clearly of interest to CEOs and often is a focus of their activities, particularly early in their tenure. As a result, reductions in corporate headquarters are often very visible decisions.

Q: You, your co-authors, and other research partners undertook a survey of 600 corporations in seven countries. Why have corporate headquarters been so under-researched? What particular assumptions were you trying to examine?

A: It is hard to get the data on headquarter sizes because companies are not required to report that level of detail. As a result, a survey is the only way to collect the data.

Corporate headquarters need to be carefully designed to fit with the strategy.

We were really trying to do five things: a) document the size and structure of headquarters, b) determine if there were any significant differences in headquarter size between countries, c) determine if there was any relationship between the size of corporate headquarters and the financial performance of the company, d) try to understand the factors that determined the size of headquarters, and e) try to produce benchmarks for companies to use in evaluating the size of their own headquarters.

Q: You wrote in the paper, "The most striking aspect of the survey results is the wide variation in the size and composition of corporate headquarters." Why is such variation possible? What does it mean?

A: There is huge variation in headquarter size basically because there is a huge variation in the underlying strategy that corporations pursue. Essentially we found that "structure follows strategy," so that the size and roles of the corporate headquarters vary with the underlying corporate strategy. A conglomerate will have a very different corporate headquarters than a firm that has diversified into a very tightly related set of businesses.

The real takeaway is that "one size does not fit all," but that corporate headquarters need to be carefully designed to fit with the strategy.

Q: What were some of the other most intriguing results from the survey? Did any of these results surprise you?

A: Perhaps the most surprising result was that the U.S. does not have smaller headquarters than European firms—indeed the opposite is true. The expectation was that U.S. headquarters were leaner than European ones because of the capital market pressures on U.S. firms. Instead, European firms' headquarters are smaller. As expected, Japanese firms had very large headquarters.

The second surprising finding was that there is no evidence that small corporate headquarters produce superior financial performance. We had expected that smaller, "lean and mean" headquarters would lead to better performance. In fact, the opposite is true—larger headquarters are associated with higher profitability. This does not mean that it is good to increase the size of a company's headquarters—it could be that profitable companies waste their money on large headquarters—but it does suggest that companies should not necessarily be looking to minimize the size of their headquarters.

Q: The lean and mean approach doesn't seem to hold water. Do you think faith in lean and mean headquarters will be difficult for managers to let go of?

A: Lean and mean will be hard to let go of, and to some extent that is appropriate. We found that there are three roles for headquarters to play: public company functions; shared services; and value adding functions.

The public company functions—tax reporting, treasury, audit, etc., that every company has to do—should be done as effectively, i.e., as lean and mean as possible. Indeed, we found it is possible to run a large 20,000-person firm with very few (perhaps twenty-seven) people in headquarters. Shared services, such as pension administration or IT, that are made available to the business units should also be subject to a market test, even if they can be quite large.

It is really only the value adding functions—corporate R&D (if the firm has such a function), business development, some HR functions, etc. —that probably should not be squeezed as tightly as possible because they do seem to be able to truly add value.

Q: Any suggestions for what our business readers should ask themselves when they look around their own headquarters?

A: Readers should look first at the corporate strategy their company pursues and ask whether the corporate headquarters size and roles fits with that strategy. They should then try and understand what goes on at headquarters in the three areas listed above. Only then can one assess whether or not the headquarters is too large, too small, doing the wrong things, the right things. If they wanted to, they could then benchmark themselves against other firms of similar sizes and industries in a very pragmatic way.

Q: How are you going to take these learnings further? What's next for you?

A: We need to finish the academic papers from this research! The real challenge going forward is that our research was primarily descriptive, using the survey data. We could not, for example, explain why European headquarters were smaller than U.S. companies headquarters—we could only document the differences. Similarly, from our data we could not resolve the question of whether larger headquarters increase profitability, or whether in fact it is more profitable firms that increase the size of headquarters. Resolving those questions is next on the agenda.

Excerpt from "The Size, Structure, and Performance of Headquarters"

by David Young, David Collis and Michael Goold

Discussion and conclusions
Corporate headquarters around the world have a wide range of sizes and structures. Nevertheless, there is a consistent set of factors that explain a large part of that variation. Prime among those determinants of the size and roles of corporate headquarters are a set of variables that reflect the firm's corporate strategy.

Corporate strategy
Nearly all the hypotheses derived from a contingent model of headquarters design, linking measurable dimensions of corporate strategy with the size and structure of corporate headquarters, were confirmed by the research. We can, therefore, assert that the design of corporate headquarters is dependent on corporate strategy.

More specifically, corporate strategies that build related portfolios will tend to have larger headquarters with stronger strategic roles. In contrast, diversity in the corporate portfolio appears to limit the role of the center. This occurs not because relatedness itself increases the size of headquarters, but because relatedness of the business portfolio leads to the selection of corporate policies that demand more headquarters staff. Companies that have related portfolios more commonly use strategic rather than financial control systems, and are more likely to influence functional activities of the business units and to actively promote linkages among those business units. All of those policy choices increase the size of the corporate headquarters. The results confirm research that has proposed that organizational design is contingent on corporate strategy (Hill et al, 1992; Goold et al, 1994; Collis and Montgomery, 1997).

Interestingly, companies with more organizational layers tend to have smaller headquarters, while increasing the span of control does not increase the size of headquarters. Only an increase in geographic scope appears to lead to an increase in corporate headquarters staffing. These findings suggest that one way to deal with diversification is to limit the role of the corporate headquarters by decentralizing more activities, either to intermediate levels of management or to the businesses themselves.

Significant industry sector differences also exist, although the major differences appear to be accounted for by a few sectors where there is potential for centralizing functional activities, particularly service activities such as IT, so increasing the number of corporate staff.

Finally, corporate strategies that create larger companies benefit from scale effects in their corporate headquarters. Typically, doubling firm size only leads to an increase of 75 percent in headquarters staff. These scale benefits are most apparent in the obligatory activities associated with sustaining the basic general management, financial and legal functions of corporate headquarters.

Non-managerial determinants
Government ownership, either past or present, is systematically associated with larger headquarters. This appears to confirm a belief that nationalization does lead to inefficiency, in this case, to larger corporate headquarters than in comparable public or private firms.

There are significant differences in the size of headquarters between countries, although similar models can be applied across all countries except Japan. Japanese headquarters appear to be larger than those in Europe, but Japanese managers also appear to be less satisfied with their headquarters and state a greater likelihood of cutting back those headquarters in the next several years. U.S. headquarters also appear to be larger than those in Europe but, in contrast to Japan, U.S. corporate managers appear to be more content with this state of affairs. One explanation is that the larger U.S. domestic market allows for more activities to be centralized at corporate headquarters than in Europe. The companion paper (Collis et al, 2002) examines this phenomenon in more detail.

Although there is a common belief that larger headquarters are less effective, we found no compelling evidence—that larger headquarters actually perform less well—particularly in financial terms. However, there is some evidence that more profitable companies have larger headquarters. However, the causal direction of the relationship between performance and headquarters size remains unresolved.

Excerpted with permission of David Collis from "The Size, Structure, and Performance of Headquarters," Harvard Business School working paper #03-096.