An Interview with C.K. Prahalad

By Joel Kurtzman

Strategy & Business: C.K., you said recently that one of management's traditional functions has been to break the company into what you call "bite-sized pieces," so that it can be correctly structured and managed. You have also said that the way this was done in the past may not be suited to the future. What do you mean by that?

C.K. PRAHALAD: These bite-sized pieces are the business unit, national organizations and so forth. They are the basic units of analysis, accountability and resource allocation.

There is, however, a problem with this. The problem is that the basic unit of analysis can be quite different for different tasks. For managing existing businesses and creating accountability, the business unit may be a good starting point. On the other hand, if you want to look at competence building, then you have to transcend individual business units. As a result, one of the really interesting questions is how to define the basic unit of analysis in a large firm. Is one definition appropriate for all tasks? The answer is probably not.

Take, for example, building a global brand franchise, or "share-of-mind" for a company. That is not the task of a single business unit. It transcends business units. Think, for a moment, of Sony. Sony's identity is consistent around the world. This identity is an important asset to protect. Are the Walkman and a broadcasting system both consistent with Sony's overall identity? Yes, they are. But you can't have Sony mean high-end, innovative products in one market, say the United States, and low-end, cheap products elsewhere. The company's identity, its skills, its reputation have to be consistent across all of its products, all of its business units, all of its markets. Is the business unit the unit of analysis in managing Sony's brand franchise? I would argue that it is not.

Now think of Canon. Canon has many, many businesses -- cameras, copiers, printers and so on. When you hear "Canon," you have certain expectations, certain assurances, promises of what the product will be like, whichever branch of the company produces it. Is the business unit the correct unit of analysis in Canon's case? Again, I would argue that it is not.

Take also competency building. It too is larger than the business unit, as are management development, skill building and relationships with partners. These issues cut across multiple business lines. If this is the case, who provides the gateway? Who shepherds those relationships?

S&B: You are not arguing that the era of the business unit is over, are you?

C.K. PRAHALAD: No. There are many aspects of managing that are still within the domain of the business unit. People have spent a lot of time thinking about how to divide companies into bite-sized pieces. They've done so for reasons that made sense in the past, not necessarily for reasons that make sense for the future. Business units were focused on accountability. They are inappropriate vehicles for corporate resource leverage -- creating skills and new business opportunities. I am saying that if we imagine the future, examine management's needs and processes, we will conclude that a total dedication to the business unit approach might be very limiting.

S&B: So, just how would you change things?

C.K. Prahalad: Let's go back to the basics for a moment. Each layer of management has its function. One of those functions can be termed "consolidation." This is really just collecting and adding up the data. It is purely a mechanical task that moves information about how the business is doing -- particularly financial information -- up the hierarchy. Then there is the job I call management's "checking-and-cost-correction" function. As part of that function, each layer checks on the levels below to make certain they are following the trajectory that is appropriate and in line with the expectations of the company or the sector as a whole. Then there are the wisdom-and-mentoring roles of management, which are to provide support to people who are on their way to higher levels of management. Finally, there is the function that we just spoke about of "getting-the-company-into-bite-sized-pieces." This enables people to track performance and it creates administrative structure.

Now, if you think about it, the consolidation role has been made very easy due to the advances in information technology. I can get data on any operation today in any number of ways. I don't need a particular level of management assigned to do the task.

In addition, the checking-and-cost-correction function -- making sure the company is on the right trajectory -- was quite important when people did not have a shared view of where the company was going. This function was particularly important when communication was poor. However, if we get people to agree to a shared agenda, so everyone knows where the company is going and how they fit in that vision, then there is little need for this layer of management either.

A good example of what I mean is quality. If we have deployed a quality program in a company, and everyone understands its importance, then we don't need thousands of people checking the quality of every product. The deployment of the program eliminates the need for the checkers. So why do we still have management checkers? The only reason we need them is that we have not yet been able to deploy strategies effectively and build a shared agenda within companies. My conclusion is that the more we are able to build a shared mindset within companies, the less we will need this administrative layer of management. This layer can become much smaller.

However, while the other layers shrink, there is going to be an increasing need for the mentoring role of management. Here the question is whether mentoring should be done only by people who have an administrative position higher than one's own, or whether it can be done by counselors and helpers who can provide mentoring even though they are not necessarily in a higher position administratively. That is an open question.

S&B: They say that Konosuke Matsushita, the founder of Japan's Matsushita Electric, viewed his driver as one of his mentors and never failed to ask him for advice.

C.K. Prahalad: Yes. Mentoring requires special skills, but it does not require hierarchical position.

S&B: Even so-called flat organizations have some hierarchical layers of management. Are these simply legacies from the past?

C.K. Prahalad: Partly they are legacies of the military system of command, and partly they are the legacy of Taylorism and so-called scientific management, where people were expected to specialize so much that a lot of coordinators were needed. It is also simply an element in the way we think. In the sciences, we tend to take complex problems, disaggregate them into their components -- or what we think are their components -- and study each component in great depth without understanding too much about the interrelationships across those components.

The goal has been to understand, in great detail, an isolated element, not the place of that element in the totality of the system. Increasingly, however, people are learning that they must understand the big picture and then the place of the elements of their specialization inside that big picture.

S&B: How do you get people to see the big picture and their place in it?

C.K. Prahalad: You train people to do this. You do it by getting people with different functional backgrounds to work together in common tasks under time pressure. You do it by getting them to achieve goals where they must understand what the other person can contribute and why the other person's contribution is important. This learning has to be experiential, not intellectual.

S&B: You have spoken about business units. What about the corporate center? Are you in favor, as so many people are, of shrinking it? Some companies, like Britain's Boots, a pharmaceutical and chemical concern, have limited the size of the center to about 100 people. Do you agree with decisions like that?

C.K. Prahalad: I am always hesitant when people come out with simple formulas for complex tasks, like saying that the corporate center should have no more than 100 people. I don't feel comfortable making a statement like that. My answer to size, structure and role questions with respect to the corporate center is that it all depends.

It all depends on the nature of the business, the complexity of the business and the tasks to be performed. For example, the Hanson Trust can run a large, multibillion-dollar portfolio of companies with only 25 to 30 people at the top. But the reason is that Hanson is a pure conglomerate. Each business has very little in common with the other businesses and the role of the corporate center is to bring fairly strict financial discipline to the companies, most of which operate in traditional, low-tech industries.

That is very different from running I.B.M., or very different from running Hewlett-Packard. In these instances, there are extraordinarily complex relationships that run across a wide range of businesses, with many shared technologies, shared distribution channels and so forth. In order to at least understand the nature of those interlinkages, and to determine what to manage and what not to manage, a very different type of top management from the people running Hanson Trust is required. The question for me is not whether we need 100 people or 200 people, but how the nature of the internal governance process changes, given the nature of the portfolio and the way in which value is created.

S&B: Are there rules by which you can make those determinations?

C.K. Prahalad: Not rules, exactly. Look at Hanson. I think of it as a "disciplined owner" that has a portfolio of stand-alone businesses. When I think of Hewlett-Packard, it is as a family of businesses, not a portfolio. HP's family of businesses has common competencies, technologies, channels and opportunities. These present different internal governance challenges for the corporate center.

Hanson's internal governance is based on financial discipline, for example. In fact, one of the vice chairmen said he had never even visited the factories of the companies that Hanson owned. He's comfortable with that and I must say I have a lot of respect for Hanson, as a company. On the other hand, if someone in senior management at Hewlett-Packard said he never visited a lab or factory or division, I would be extremely worried.

For these reasons, I have great difficulty in understanding why people look at a company like General Electric, for example, and say we should be like that. G.E. is a unique entity. There is only one G.E.

As you can see, the question of the corporate center has a different answer, depending upon who is asking it.

S&B: You mentioned internal governance. How do you see that working?

C.K. Prahalad: I have been giving a great deal of thought to a number of closely related ideas. These have to do with strategy and with the processes of governance within companies. I'm less worried about the external governance issue, which is a topic of debate among shareholders, boards and C.E.O.'s. By internal governance, I mean the ways in which things are actually accomplished -- or not accomplished -- inside companies. It also has to do with who does those things and for what reasons.

The focus of my analysis is the internal governance process. By the time the C.E.O. sees a negative result in the P&L statement, in market-share loss statistics or in the company's inability to deliver new products quickly, it's simply too late. The cancer has already set in and large companies have an enormous amount of inertia when it comes to change. That means the real issue is how do you prevent the cancer from developing in the first place? How do you develop a methodology for assessing the quality of internal governance, especially in a multi-business company?

I think that question raises interesting issues about the nature of the relationship between headquarters and subsidiaries -- which is an old problem -- but it does so in a new context. It also raises interesting questions about the future and who should think about it.

S&B: What is the new context to which you are referring?

C.K. Prahalad: It is simply that today you cannot allow subsidiaries to be totally autonomous -- nor can you expect them to be slaves of headquarters. That presents a tension.

S&B: Is the debate, therefore, about autonomy versus control? Is it, as Charles Handy maintains, about subsidiarity? Or is it about centralization and decentralization?

C.K. Prahalad:
I am afraid it is about neither of these. My starting assumption is that the debate gets framed as centralization versus decentralization. But that is the wrong way to look at it. I think the debate ought to be framed in terms of the value that is added by each layer of management. The debate should be about how various groups -- staff groups and line groups -- at various levels and in various places in the organization really add value to customers and add value to the people who have to deliver a functionality to the customer. The debate should not be about power, nor should it be about egos, span of control, decentralization versus centralization or size. The debate should be about adding value throughout the organization.

If that is the case, then the question really becomes how do you create structures internally so that relationships are based on value creation. If you look at structure from a value-adding perspective, then there is an important issue related to internal governance that must be addressed. I frame it this way: Who has responsibility for competitiveness?

Almost all traditional views of strategy have been very elitist when it comes to this question. The assumption is that top managers develop strategy and everybody else implements it. I don't think that is the correct answer to the internal governance question.

S&B: What is the correct answer?

C.K. Prahalad: Let me say the following before I get to my answer. We are in an era of discontinuous change, whether we are talking about telecommunications, health care, financial services, high-volume electronics, retailing or the Internet. As a result, we are no longer talking about fine tuning or improving the organization's efficiency. We are talking about nothing less than reinvention -- reinventing the business in fundamental ways.

Reinvention requires a new skill mix and new ways of approaching the business. It may require different business models. It may also require different people. It is not at all clear to me that the same people who are socialized in the standard way of doing business -- and who understand a certain recipe for how to manage -- can change very quickly and become the inventors of the new business model. As a result, companies have been looking outside their basic industries for people who can thrive in the new environment. Look at George M.C. Fisher -- he came from Motorola to run Kodak. Or look at Louis V. Gerstner Jr. -- he came from RJR/Nabisco to run I.B.M. They were hired to bring a different perspective to these companies.

Now, to your answer. I have been experimenting with ways to enable people lower down in the organization -- people who are closer to new technologies, to customers and to competitors -- to create the point of view and dialogue that is needed as a prerequisite for change.

S&B: How has it worked?

C.K. Prahalad: So far, where we have tried it, the results have been successful. We have found people deep down in the organization to be receptive to change. We have also found tremendous creativity and knowledge there -- so much so that senior managers have been quite surprised.