What Does It Take to Compete in a Flat World?Published: October 31, 2007 in Knowledge@Wharton
When Thomas Friedman wrote his popular book, The World Is Flat, one of its central arguments was that geography might soon become history. The proliferation of information technology and telecommunications networks has integrated the world in ways that were unimaginable in the past -- and this has transformed how companies produce and distribute products and services. One result of this transformation is the rise of networks of companies that are bound together through IT and logistics. How can firms strive for and gain competitive advantage in such an environment? Victor and William Fung, group chairman and managing director of Hong Kong-based Li & Fung, and Yoram (Jerry) Wind, a professor of marketing at Wharton, deal with this issue in their new book, Competing in a Flat World: Building Enterprises for a Borderless World. They recently spoke with Knowledge@Wharton.
An edited transcript follows:
A Note to Readers:
Knowledge@Wharton and Wharton School Publishing are pleased to announce a global competition for the best strategies for competing in a flat world. Two prizes will be awarded, one for the best strategies that work and one for the best insights from failures.
Knowledge@Wharton: Your book, Competing in a Flat World, describes Li & Fung as "a flat business for a flat world" and also argues that you have evolved into a "network orchestrater." What does that mean and what implications does that have for your customers?
Victor Fung: Maybe the way to explain this is to start off by describing the fact that the way manufacturing is done today -- versus, let's say, even as recently as 10 years ago -- has fundamentally changed. There has been a major transformation over the last 20 years, culminating in the recent period, because of the presence and the use of enablers like IT and logistics.
Whereas manufacturing used to be done in one factory, under one roof and in one country, today the manufacturing process is being dispersed and being produced in stages, in many factories, sometimes over many countries. So when you have a particular order -- let's say of 100,000 shirts -- to manufacture, you may in fact do different stages of the production process in different factories and in different countries. Then you put everything together to make the final product using IT and logistics.
As a result of this change, when we make any product, we have to orchestrate a network of processes and suppliers in different parts of the world. It is very important for us to work very closely with this network; we don't control the entire network because they are all independent contractors. Therefore, when we talk about "orchestrating" that network, it means we have to make sure that all the pieces are synchronized and that the production comes out with a high quality -- and also with a very fast turnaround time. That would allow us to be very efficient in the production process.
Knowledge@Wharton: William, would you like to add anything to that explanation?
William Fung: Well, as Tom Friedman says in his book, the world is getting flatter in the sense that it is now easier to communicate and for parts and materials to go from one country to another for processing. Our ability to use the best factories, at the best locations, to manufacture any order, is becoming a reality.
Instead of the factory being located close to the market where its goods are consumed, production of a lot of labor-intensive goods can now take place in countries where labor is plentiful rather than in countries where these goods are consumed. That accounts for the necessity of a business like Li & Fung -- to be able to operate without borders in a flat world.
We have organized the business so that we have a very flat organization to cope with the need to respond quickly to the needs of our customers and without a hierarchical structure. That's the flat business aspect of the way we're organized to operate in this flat world.
Knowledge@Wharton: Li & Fung has a business history of more than 100 years; you started out in 1906 during the Qing Dynasty. How did you evolve into a network orchestrater?
Victor Fung: If I could take a stab at that question...a lot of the examples and the experiences that we cited in the book were because of the pressures of customer and consumer demands, rather than just some smart people saying that this is the way that we should be organizing the business. The business was organized in a way that responded to the needs of the market.
The response to these trends of the needs of the market started way back...I would say after the Second World War. This was when the prosperity in the developed markets, like America and Europe, led to a situation where a lot of the young people were unwilling to take up blue-collar jobs in labor-intensive industries to make the products that were being consumed in those markets.
This rationalization of manufacturing of labor-intensive goods around the world started in the 1960s and 1970s. Originally Hong Kong was one of those places where our whole economy, after 1949, became dependent on manufacturing for export and not for home consumption. This ability to manufacture in a different location from where you consume it started to spread from Hong Kong to places like the Asian Tigers [Taiwan, Singapore, Hong Kong and South Korea] and eventually to Southeast Asia. This rationalization of manufacturing led to a situation where we needed to create the kinds of supply chains that oversaw the manufacturing of these labor intensive goods and eventually to bring them back to the markets of America and Western Europe, where they were consumed.
We were reacting to this economic situation -- the ability to manufacture at less cost in countries where labor is plentiful and then to bring goods back to the countries where they are finally consumed. As a result, we had to organize the company in a way that we could control these more complex global supply chains.
William Fung: Today we are responding to a very important consumer trend. There has been a tremendous fragmentation of consumer demand. Instead of serving one huge market that is basically uniform in demand, you are seeing pockets of niche markets. That is becoming more and more evident.
As we see these niche markets develop, there is a need to come up with increasingly different products. At the same time, forecasting demand becomes much more difficult. This is because if you break up a large market into a fragmented number of smaller markets, the amount of wiggling around in each of the markets becomes more severe.
Therefore, in order to buy more accurately the right product at the right time -- and to deliver it at the right time to satisfy the demand -- you need to cut down the re-order cycle. We talk a lot today about "fast-response manufacturing." This idea of dispersed manufacturing, that I described earlier, allows us to use triple sourcing and double sourcing at different stages. This will give us a very fast turnaround time and the ability to shrink that turnaround time. As a result, we can delay the ordering as much as possible to satisfy the market demand as long as we can. In effect, it allows us to buy more accurately.
The reason this is all possible, of course, is because in the last 10 years there has been a tremendous development of information technology, the Internet and also what I would call modern logistics. With those two enablers we are able now to carry this model of dispersed manufacturing.
We are taking advantage of all of the factors that are described in our book Competing in a Flat World to satisfy this consumer trend towards more and more specialized market niches and different categories of products to satisfy different groups of consumers that we see today. I can see that this trend is actually just going to continue further and further as we go forward.
Knowledge@Wharton: Jerry, from your perspective how does the Li & Fung model of network orchestration compare with the models of other companies that are trying to deal with the same phenomenon of doing business in a borderless world?
Wind: We have to distinguish here between two situations. There is the immediate parallel -- other manufacturing companies today are concerned with supply chains. Every company today is struggling with how to come up with the best supply chain, and there are different models. Toyota, for example, has a dramatically different model with much closer links to their suppliers. This is different than the situation with Li & Fung, where they have close to 9,000 suppliers around the world, in many countries and they have designed a custom supply chain from this network.
The challenge and the strength of the Li & Fung model is selecting and creating this network of 9,000 or so suppliers -- and then for each order to customize a custom supply chain that is the best in delivering the right product, at the right price, at the right time, based on the right specifications of the consumers. Obviously when you're dealing with the supply chain area, the lessons from Li & Fung are tremendous. This is terms of how to do it more effectively and how to create the best possible offering for each one of their retailers and therefore meet the needs of the customers, as both Victor and William articulated before.
But the network is not limited to manufacturing supply chains. We see networks all over, in a sense. In every aspect of life today, we are encountering networks and there is no company today that can survive as an island, as a company by itself. That's why there are the companies that deal with networks of partners and others.
At the other extreme, consider Wikipedia, the free online encyclopedia -- which has been receiving a lot of attention. Wikipedia can be viewed as a network of participants; but there too it's not that it's totally organic and there is no governance and direction -- there is an orchestrater. There is a governing body and there are rules. So even the totally open Wikipedia follows some type of network orchestration guidelines that specify that you cannot alter certain entries or certain rules implied and the like.
Another example that people are familiar with is eBay. The company provides the platform and in a sense is orchestrating buyers and the sellers using its technology. The lessons from the network orchestration that Victor and William have created at Li & Fung are applicable to a wide range of businesses and not only on the manufacturing side.
Even those who are not involved directly in business may find it interesting to understand how value is created into today's environment - to meet the changing needs of customers - in this changing flat world.
Knowledge@Wharton: What are the major challenges involved in network orchestration and how do you tackle them?
William Fung: The major challenge we faced is the fact that the needs of the market are changing very rapidly. The need for flexibility is the reason why we went into a network of suppliers instead of a single monolithic, sort of hardwired, vertically integrated manufacturer in the first place. The need to keep providing all of the options needed by the market is obviously always a challenge.
Now couple that with the changing economics of the manufacturing in different parts of the world, and you will see that the network itself - it's a moveable feast. It is not something where you can say, "I've built now, and this network will last me for a long time." It doesn't work that way - it keeps evolving. Our challenge is constantly finding the right countries, the right factories within those countries, to provide the economics that's needed and sometimes the speed to market or the efficiencies needed by the market.
So one of the major challenges of this network creation is that it is ever-changing and we always have to be updating and renewing it. A lot of the products we are involved with are labor-intensive types of products to manufacture. And with labor intensive industries, it's almost like the first rung of the industrialization ladder for developing countries that are changing from an agricultural to an industrial economy.
Therefore our challenge has been to keep exploring new frontiers, new countries to do our manufacturing in that provide a better economic or other solution. It's not purely economic; it could also be sometimes that these countries have the raw materials that are needed for any particular product...wood for furniture, cotton or silk for our garments, or something like that. It's always an evolving situation and we constantly have to be on top of it.
Victor Fung: One of the main things you should know is that once you embark on a dispersed manufacturing model, where different parts of a manufacturing process can be done in different factories and in different countries, for a country to be participating in a global economy it no longer has to have vertical integration of the whole industry. In other words, you don't need to go from yarn to weaving and then making the garment to be in the garment industry. If you are just in one segment of it, you can participate in that segment of the total supply chain.
That allows for a number of phenomena. First it allows the opportunity for much larger and wider participation for small and medium sized businesses globally. This, I think, is a huge phenomenon that is happening in the world today with a lot of consequences for economic development and job creation.
The second is, for countries to compete, they no longer have to build an entire vertical before they become competitive. They can actually be in a particular niche in a vertical. And indeed, they may be very specialized in a particular niche and then they can participate in different verticals. So it's an interesting phenomenon that's happening because this model is being implemented so widely now around the world.
William Fung: If I could add to that, with another example, with an industry that we're not actively involved with, like the automobile industry. Auto parts are now produced in different parts of the world. If you look at something like, for example, the small electric motors that power windows in a car in let's say Germany. These are now made primarily in the Far East, either in Japan or in China. There are so many well known companies that have specialized in these niche markets as Victor has said, of creating these products and trade and have basically become world leaders in that particular part or component. And, I think this is what the network of dispersed manufacturing can do for the industries, rather than having to do everything yourself in a vertical set-up.
Knowledge@Wharton: What difference does it make if your company is an orchestrater or it's orchestrated by someone else?
Victor Fung: What we are talking about is team work. Whereas before, you might say that in the old model of a hierarchical, vertically integrated company, everything happened within one company, now we are not seeing competition between companies, but between different supply chains. In a sense, every time you see a product competing, it's one supply chain against another supply chain, with a number of players involved.
So once you have a team, there is always a need for somebody to orchestrate the team. Being a team member [I think] is a very important role. And then, for companies like Li & Fung, where we tend to not be doing any of the actual manufacturing ourselves, but we orchestrate the entire process of allocating the work, sometimes designing the product, getting the orders and so on. We act as an orchestrater -- and that is I think the major difference. On a full supply chain you need both types of players in order to be very strong.
William Fung: Yes, because if you don't have an orchestrater, what tends to happen is that people will focus on their particular sector of that supply chain and they sub-optimize. You need somebody to look at the holistic needs. A very good example would be that if you need the supply chain to respond very quickly to changes in consumer demand, rather than have a sort of buyer-seller relationship that is up and down your chain, you need a cooperative relationship.
This is where you say that "We don't know what the ultimate demand will be, but let's keep open capacity, so that when the orders do come, we can respond very quickly." If every player in the supply chain only minds his own sector, you're not going to have that kind of flexibility built in. And in order to do that, you will need someone to look at the whole chain to see how it is responding, and for them to say, "We need flexibilities built into this part of the chain and that part of the chain." Sometimes it's the raw materials supply; sometimes it's in making certain components and sometimes it's in the final assembly.
That is the difference between an orchestrater and a player, or someone who is part of the chain. You need someone to really manage the whole chain, as Victor has talked about.
Wind: But, if you talk about this from the point of view of management in general, at any firm, then I think that in every firm, the top management should consider what role they can play and if they can start playing the orchestrater role. So, this is going beyond the supply chain and the manufacturing side, with any firm that you are dealing with, whether it's in the financial services area or others - you have to basically look at the entire operation and try to be the orchestrater.
So in a sense, you can argue that every firm makes or boasts of an orchestrater and is also part of an orchestra - part of a network. But, ideally in terms of the changing management scene, we really see it as the role of management -- when they start thinking about division of the firm and the strategy of the firm. They need to start looking at this more broadly and say "How can I orchestrate the network that I am in charge of and what is the best way to do it"? They need to really perform more in the role of an orchestrater.
Knowledge@Wharton: Assume that you are the orchestrater. How should you structure incentives and penalties for the network members? In other words, how do you balance control and empowerment within the supply chain?
Victor Fung: That's a very good question. We have about 9,000 suppliers worldwide, operating in more than 40 countries, so it's a very large global network. And we keep in touch with that network through a very sophisticated IT system. Now, for every individual supplier, or member that we work with, our objective is take anywhere from 30% to 70% of that particular factory's output. You might ask, why 30%? Well, 30% is because we want to be large enough, as a potential buyer, to be meaningful to the factory. But, on the other hand, we don't want to take much more than 70%. This is mainly because we want that factory to be able to work with other people outside of the network, in order to get new ideas, new techniques, and so on - so that the network is constantly being renewed.
You can see that within the network itself, we would then have really, what I would call a loosely orchestrated network - in which people are not participating 100% of the time in the network - they could be participating outside. And some people go in and out of the network. I'd like to describe the state of that network first.
Now, in terms of structuring the incentives, there are two major ones. The most important, is that being part of the network, we would constantly be feeding ideas, feeding orders, feeding new product development trends and so on, to the members of the network. And, we expect the network members to also contribute because they have exposure to people outside of the network.
The second, very important aspect is that Li & Fung has a whole spectrum of customers, going from buyers of low-cost and value-oriented basic products to very sophisticated luxury products. For any particular manufacturer, when they are operating at a certain level... we don't keep going back to the manufacturer every year and saying "We want you to keep producing the same thing, but at a lower cost every year." Instead, we try, over time, to migrate that particular member of the network up the value chain -- as we can keep increasing the number of sophisticated customers and products for that particular factory to work on.
One of the big incentives is this idea of being able to learn from us as an orchestrater, to be able to depend on us for a reasonable amount of the work that the factory does - and most importantly having the ability to use us in the network to upgrade oneself to actually higher and higher value added and more and more sophistication. I think that is very important. That, to me, is the real incentive for a member to stay as an integral member of the network.
Knowledge@Wharton: Is what you were describing what you call the "30-70" rule in your book?
Victor Fung: That's right. That is the way we conceptualize a fundamental member of the network and that gives you an idea of how the whole network orchestration happens with these 9,000-plus factories, in more than 40 countries around the world, which form a part of our network.
Knowledge@Wharton: How do you decide who gets to be in the network and how long they can stay in the network?
William Fung: That is dictated by the needs of our customers. We have a wide range and a large number of customers, as Victor had pointed out, and different levels of requirements. But generally speaking, I would say that we select our factories very carefully. But even after the selection, it's through a process of working with them on specific orders and specific customers that these companies started to evolve [as members of the network].
Although Victor has said that IT is very important -- and it is -- we also have a lot people in the network who make qualitative decisions about how factories perform and how they are doing. The network itself is manned by 12,000 dedicated people. We don't have anybody who is manufacturing per se. We only have people who are involved in network orchestration.
We have 12,000 people in more than 70 offices, in more than 40 countries, around the world. It is this network of people who are scrutinizing, improving and changing the network of supplies all the time, depending on both the requirements of the market and the performance of the supplies themselves.
Knowledge@Wharton: What are the principal political and policy implications from the rise of these networks? For example, do you think that there can be regulatory oversight in the context of nation states of such networks?
Wind: This is a topic that has been very close to Victor, for many years. This is an area where traditional trade statistics don't make any sense. If a product is manufactured in four countries, or six countries, what is the country of origin? I think this requires a fundamental change in the mental model that we have as to what is the country of origin because the original concept no longer makes sense... So, I'll let Victor answer.
Victor Fung: My view is very simply this: If you think about the WTO and the way we measure trade flows around the world, it assumes a very simple model, in which a product is made in Country A and it is shipped to Country B to be consumed. That is the fundamental assumption of the model. The question is where does the substantive transformation that creates the product take place...because that is where all of the value-added is.
But if you look at the model of dispersed manufacturing, the same product actually can be made in several countries before it becomes a finished product. And, if you say the substantive transformation then occurs in the final stage of finishing and basically that becomes the country of origin -- that last country, so to speak, gets charged for the full trade statistics, whereas it may capture only a portion of the total value added, it may only capture say 30% of the value added.
And so the problem with traditional trade statistics is that you are using a system to describe a world that has already evolved and no longer fits the model. Therefore, you get all of these distortions. A case in point is, for example, you hear about this huge trade surplus between China and the United States and it is causing a lot of friction in terms of trade relations and so on. But if you really look into it, maybe only 30% of the value-added occurs in China. A lot of the stuff is actually produced in the initial parts of the supply chain which originate somewhere in Southeast Asia. The parts and materials may go through one or two countries, before they end up in a product that is finished in China - but China gets charged for the full trade statistics.
So one of the most serious policy implications, if this model gets more and more widely used, is that the whole way that we look at the country of origin, the definition of where a product becomes substantively transformed, and the way we keep trade statistics, needs to be reexamined.
William Fung: Let me illustrate what Victor has said with a concrete example. If you look at your laptop, chances are that the memory device could have been made in Malaysia, the monitor could be made in Taiwan, and the laptop itself might have been assembled in China. But the highest value might be the Intel inside, the CPU chip.
Under the present trade statistics rules, if China exports that finished laptop to the United States, that becomes a Chinese export. Therefore, the trade flow is defined that way and it appears to be a Chinese export. But the value added on China's part, which is only the final assembly - which uses a lot of fairly cheap labor processes - could be far less than all of the other components entered and the technical software that goes into making that laptop.
So, you can see how the trade statistics could be totally distorted if you look at it in terms of value added -- rather than the final place where, as Victor said, it has been assembled. And, I think that there is a complete revamping needed in the world on how you record trade flows. This is because too much emphasis has been put on trade deficits leading to a call for sanctions against trade partners and so on.
And, I think that there is a rather unnecessary or almost impossible to achieve need for bilateral trade balances. You know, it's almost impossible under this new world, where many countries participate in making a product... to call for bilateral trade balances between two countries -- whereas many countries participate in this whole process.
Knowledge@Wharton: What kind of policy changes would you like to see to address this situation?
Victor Fung: Well, we need to re-examine the whole issue of country of origin and not capture trade statistics the way that we are doing it. It should depend on allocating some concept of value added to different countries. But, of course this is a very deep theoretical issue that needs to be examined.
I'd like to touch on another major policy implication. As we see this model of dispersed manufacturing emerging further and further, what I see is an opportunity for more and more participation from small- and medium-sized enterprises from around the world.
At Li & Fung, because we splice the supply chain so finely and optimize each portion, we keep on talking about atomization of the supply chain. That is sort of an in-house term. But I think that the atomization of the supply chain also in a sense means the democratization of the supply chain since it allows more SMEs [small and medium enterprises] and smaller companies to participate. Many, many of our 9,000 suppliers from around the world are small and medium size enterprises. And, I think that that actually provides the world with a model for economic development, especially for some of the developing countries which is very crucial. It allows even SMEs in developing countries to participate in global supply chains.
Now, of course, there is a minimum price of entry. And in my mind, that minimum price of entry is that the governments that host these SMEs need to at least provide the SMEs with the minimum level of access on the IT front -- so that they can participate. And, also they have to have the infrastructure to be able to deal with the logistics of the goods flow in a very fast response manner. But once the government can actually provide those two enablers of IT and infrastructure logistics, then I think that SMEs can really participate. This democratization of the supply chain is a major trend in the world -- a major phenomenon and I think it is a very good development.
Knowledge@Wharton: One final question for all of you: What major lessons have you learned through this process of network orchestration that you'd like to share with Knowledge@Wharton's readers and listeners around the world?
Wind: The key lessons are captured by the framework that we propose in the book of trying to deal with network orchestration. They are primarily composed of three major dimensions. One of them is the need for a firm to start thinking not just about itself, the traditional firm - but how to balance the interest of the firm with the interest of the network.
I mention the concept that Victor had talked about in terms of competing network against network as opposed to firm against firm -- which has huge implications. So, how do you create this balance, realizing that you are operating as part of a network?
The second dimension we highlight is how you move away from traditional control. This is because in reality Li & Fung does not own any of the 9,000-plus companies. But yet, it has to deal with empowered companies, empowered consumers - so there's the whole balancing again between having a new type of control and the empowerment of the participants.
The third area is the need for balance between the traditional focus on specialization and capabilities for integration. It seems what we have here is a new paradigm for management that focuses on the network and the need to balance the network in the firm. This leads to new capabilities, new types of competencies that we need in terms of the orchestration of a network. So, we are learning from the lessons of this project and we're extremely excited about some of the implications. We are having a conference that is coming up in November and it is on network based strategies and competencies. This has been stimulated to a large extent by the work of this book.
William Fung: From the standpoint of running a firm on a day-to -day basis, one of the key words in my mind is speed and fast response. Your whole being is oriented now towards speedier response. The other word is outsourcing...radical outsourcing and orchestration of a network in that you don't really control everything in-house.
But at the same time you're focusing completely on turnaround time. If I were to construct a factory today, for example, the way that I would put a factory together would not be how to make very efficient long production runs. That is a thing of the past. The key in my mind is how to how to satisfy customers with very low minimum order quantities. You want to arrange a factory so that you can make line changes, very, very rapidly. And that flexible sort of production is going to be the key to the future.
The way that you run an organization lightly and perform as a network orchestrater is that I think you really have to think constantly about speed of response. This means a very flat organization, total empowerment for people on the front lines, so that they can make decisions instantaneously on the spot. Also, you need to have a very strong IT network so that they have this information at their fingertips. At the same time, you have to think about an incentives system that will allow them to feel like they are entrepreneurs and that they are actually running their own small company. This has very serious implications.
If you can think of the opposite paradigm -- that the world did not go this way of global supply chains and orchestration of global networks but it went the other way with vertical integration -- then you would see a smaller and smaller number of larger and larger vertically integrated companies. Indeed, at one stage of the development, maybe 30 or 20 years ago, you actually did have this idea of saying that everything should be vertically integrated in house. That would be a completely different model. That would be a very vertically oriented hierarchical model.
And yet, we now have gone to a radically outsourced, very flat and very fast response organization. So, I think that that's the contrast and to me that's the major lesson from a managing of the firm standpoint.
Knowledge@Wharton: That's interesting. Would it be correct to assume that in your model network orchestration is not monopolistic? Or does the network become the monopolist?
Victor Fung: As I said earlier, it is the democratization of the supply chain. In fact, to us it's "the more the merrier. The more that we can get people into the network, the more we can actually have a network that hooks together more and more supplies -- the more opportunities we will have to optimize -- and the more we can involve people from different parts of the world and then they can bring different ideas -- a way of operating, and different capabilities into the network.
Wind: It is extremely important that you keep in mind that the lessons are really at all levels. The levels are obviously to the firm. The lessons are, in a much larger context, to public policy around the world and for economic development. And the concept of network orchestration is a very, very powerful one. So that even though there are implications in every one of these areas -- there are even implications for us at Wharton and in business schools around the world - in terms of what is management. Management under the model of network orchestration is dramatically different than the traditional concepts of management.
A Note to Readers:
Knowledge@Wharton and Wharton School Publishing are pleased to announce a global competition for the best strategies for competing in a flat world. Two prizes will be awarded, one for the best strategies that work and one for the best insights from failures.