4

LEVERAGING CHINA AND INDIA FOR GLOBAL ADVANTAGE

In five to ten years from now, when we think of AMD, we will think of it as an Indian company, a Chinese company, and a North American company. The three regions will cover all the crucial areas, namely: Where the products come from, where the technologies are developed and where the most advanced markets exist.¹

Dirk Meyer, president and chief operating officer, AMD

The Qinghai-Tibet Railway, which started operations in July 2006, is the first time in history that the Tibet Autonomous Region has been connected by rail links with other parts of China. It is also the world's highest-plateau railroad running on the “roof of the world,” with about 960 kilometers of the track located 4,000 meters above sea level, the highest point being 5,072 meters above sea level. GE supplied the locomotives for these trains. The bulk of the engineering work for these locomotives was done at the company's John F. Welch Technology Centre in Bangalore, India, with guidance from senior engineers in Erie, Pennsylvania.²

Like many of its other products and services, Apple's iPod transformed the music distribution business. It was also the first major product Apple introduced after the Macintosh and its various versions. Although Apple's CEO Steve Jobs and his colleagues led the overall design, the bulk of the work on the iPod was done in India and China. Apple outsourced the “brains” of the iPod (a microprocessor) to Portal Player, a Silicon Valley-based semiconductor company. Portal Player's engineers in Hyderabad, India, and Silicon Valley worked around the clock to design the chip. The chip itself was manufactured in Taiwan. The final assembly of the iPod took place in China.

In December 2007, GlaxoSmithKline, the U.K.-based pharmaceutical company, announced that it would invest over $100 million over the next twelve months to build a major neuroscience research center in Shanghai, China. Once up and running, the center would be responsible for virtually all of the company's research on neurodegenerative diseases. Explaining this decision, Moncef Slaoui, head of the company's R&D operations, noted, “For us, China is not about outsourcing and cheap labor. We don't want to give them the crumbs. It's about different science. We will link our fate to their fate. Within five to ten years, we will be moving from ‘made in China’ to ‘discovered in China.’”³

In late 2007, Cessna Aircraft Company, a subsidiary of U.S.-headquartered Textron, announced that it will outsource the complete production of its Cessna 162 SkyCatcher model to China's Shenyang Aircraft Corp. Jack Pelton, Cessna's president, explained that without partnering with Shenyang, Cessna probably would not have started the SkyCatcher program. According to Lewis Campbell, Textron's chairman and chief executive, outsourcing to Shenyang would enable Cessna to sell the planes for about $109,500 each: $71,000 less than what it would if the planes were built at the company's factories in Kansas.⁴

As these examples illustrate, the comparative advantage of China and India has broadened considerably beyond cheap labor to also include leading-edge talent in some of the world's leading-edge industries, as well as home-grown innovation in technologies, products, processes, and even business models. In this chapter, we begin by outlining the potential opportunities for competitive advantage that China and India offer to any multinational enterprise. We then examine what the enterprise must do in order to convert the potential into reality.

A Look at the Opportunities

There are three primary dimensions along which China and India are becoming central to global competitive advantage for a rapidly growing number of companies across a wide range of industries: cost arbitrage, talent arbitrage, and innovation.

Each of the three sources of competitive advantage can be hugely important on its own. However, if they can be leveraged in tandem, the impact can be especially powerful. The biopharmaceuticals industry shows how. Industry estimates are that for every drug approved for sale by the U.S. Food and Drug Administration (FDA), the process begins with an analysis of 5,000 to 10,000 compounds. Of these, about 250 enter preclinical testing. Only 5 of these make it to clinical testing, from which 1 may finally pass all hurdles and get FDA approval. The process can take up to fifteen years and cost $1 billion. Consider now that both China and India produce five times as many chemists as the U.S. does at the bachelor's level and three times as many at the master's level. Furthermore, depending on qualifications and city location, this talent costs only about one-fifth to one-third of that in the United States. In short, for the same or lower R&D budget, a company such as AstraZeneca or Novartis can hire a much larger number of analysts in China and India and significantly accelerate the pace of drug discovery.

The other side of drug development is clinical testing. Given the large populations and low income levels in China and India, enrollment in clinical trials can be fast, easy, and highly efficient since a single site can recruit a much larger number of patients. Low income levels also imply that it can be relatively easy to find large numbers of treatment-naive patients (patients who have never undergone any treatment for the particular disease), thereby greatly reducing the likelihood that the results from clinical trials may be confounded by the effects of previous drugs and thus improving their reliability. On top of the speed and efficiency of patient recruitment, India (more so than China at present) also offers robust skills in clinical data management, which can further help reduce the time lag in making sense of the results from clinical testing. To sum up, the powerful combination of lower cost and a vast talent pool can be leveraged by a global pharmaceutical company to accelerate the pace of product innovation significantly.

The same logic applies to an increasing number of industries, from semiconductors to cars to medical diagnostics. AMD is a microprocessor company and archrival to Intel. Survival and success in this industry depends crucially, although not solely, on the scale, effectiveness, and efficiency of R&D activities. However, given brutal industry rivalry and losses over the past two years, AMD can ill afford to increase its R&D budget as a proportion of sales revenue. What it can do, however, is to try and get much more bang for its buck by increasing the scale of R&D operations in places such as China and India, where the size of the available talent pool is as large as or larger than in the United States or Europe and the costs much lower. We see this as one of the primary reasons that Dirk Meyer, the company's president, can convincing say that, with each passing year, his company will be not just an American company but will also become an Indian company and a Chinese company.

We now take a closer look at each of the three sources of competitive advantage.

Cost Arbitrage

Notwithstanding ongoing escalation in Chinese and Indian wages over the past several years, labor costs for most types of blue- and white-collar work in both countries remain well below those in the developed economies. Given that per capita income in China is about one-twentieth and in India about one-fortieth of that in the United States, this is exactly what we would expect. Table 4.1 provides illustrative data from two companies in our field research.

Table 4.1 Labor Cost Comparisons, December 2007

Note: Estimated total costs, including benefits.

Source: Data from the authors' field interviews, December 2007-April 2008.

As a broad generalization, these as well as similar data from other companies lead to the following conclusions about the reality as of December 2007:

• For blue-collar work, costs in the United States are about fifteen to twenty times higher than in China or India.
• For junior engineers (age range twenty-two to twenty-five years), the United States is about nine or ten times more expensive than China or India.
• For more experienced engineers (age range twenty-seven to thirty years), the United States is about five times more expensive than China or India.
• For department heads (age range thirty-five to forty-five years), costs in the United States are about three times higher than in China or India.
• For senior executives (such as company president or other C-level executives) of similar-sized companies, cost differences between the United States, China, and India have narrowed considerably or even vanished entirely.

These cost differences provide considerable opportunities for cost arbitrage for most companies in most industries. It is precisely for these reasons that Sony has shifted virtually all production of consumer products from Japan to China and why IBM has transformed its services business into a heavily India-led global enterprise.

Talent Arbitrage

Measured in terms of the number of Nobel Prize winners and citation impact of published journal articles, the United States continues to reign supreme in terms of the quality and scale of the scientific research being performed within its universities. In the scientific fields, there is no living Nobel Prize winner who did his or her pioneering work at a university or laboratory in China or India. Although these two countries are getting stronger in pure scientific research, it will take at least ten years before they can have a notable impact on scientific knowledge.

Table 4.2 Bachelor's Degree Graduates in Engineering, Computer Science, and Information Technology (in thousands)

Source: Abstracted from V. Wadhwa, G. Gereffi, B. Rissing, and R. Ong, “Where the Engineers Are,” Issues in Science and Technology, National Academy of Sciences, Spring 2007.

Notwithstanding the salience of pure science, it is important to note, however, that more than 99 percent of the talent that powers economic development (at the country level) and competitive advantage (at the company level) is of the relatively more mundane kind, such as chemists, computer programmers, clinical research assistants, laboratory technicians, engineers, statisticians, financial analysts, and architects. For these types of skills, the quality of education at many colleges in China and India is not only comparable to that in the West but is getting better. These two countries also offer the advantages of scale and lower cost.

Table 4.2 provides data on the annual output of bachelor's degree holders in engineering, computer science, and information technology in the United States, China, and India.

As these data indicate, China and India already produce a larger number of engineering, computer science, and information technology graduates than the United States. Furthermore, enrollments in China and India are rising rapidly, while those in the United States are fairly stable. Thus, it is all but certain that the gap in the size of talent pool will increase rapidly over the next ten years. Similar trends are likely at the master's-degree level. The current output of master's-level graduates in computer science and engineering is about sixty thousand in each of the three countries. However, the numbers in China and India are rising, while those in the United States are flat.⁵ China and India will also benefit from the growing proportion of foreign-trained scientists and engineers who are returning to their native lands.

The scale, quality, diversity, and cost of the available and growing talent pool in China and India offer significant opportunities to a global enterprise that knows how to leverage this talent for global advantage. Over the past ten years, an increasingly large number of companies have already figured out how to leverage Indian talent for activities such as call centers and data entry. In addition to these well-recognized opportunities, a whole spectrum of new opportunities is now emerging in more knowledge-intensive domains such as product and process R&D; analysis of complex financial deals and instruments; legal research to help a company deal with, say, patent or merger-related issues in American or European courts; mining of market research data to derive strategic implications; back-end engineering work on architectural projects; and conversion of books, annual reports, and technical manuals into digital formats so they can be accessed over devices such as a mobile phone or a personal digital assistant, to name just a few. Given that virtually all college-educated Indians have a very high level of fluency in the English language, India continues to enjoy a significant advantage over China in many of these tasks. Until China closes the language gap (which, in our judgment, will take at least ten years), its location advantage is likely to be primarily in research and development, where it leads India and where daily or frequent communication with peers and customers in other countries is generally not necessary.

Leveraging for Innovation

Given their vast talent pools, China and India are already playing an increasingly important role in technological and product innovation. This is evident from Table 4.3 which contains trend data on the number of patents granted by the U.S. Patent and Trademark Office to all inventors, foreign inventors, and inventors based in China and India.

Table 4.3 Number of Patents Granted by the U.S. Patent and Trademark Office

Source: U.S. Patent and Trademark Office: http://www.uspto.gov/go/taf/cst_all.htm.

While the absolute numbers of patents granted by the U.S. Patent and Trademark Office to inventors based in China and India are small, note that the data pertain to patents granted rather than patent applications filed; the latter figures are almost certainly much larger. Also, the rate of growth in patents granted to China- and India-based inventors is very high. Thus, by 2015 to 2020, their share is likely to have grown significantly. It is clear that for any company whose competitive advantage depends significantly on R&D, ignoring China or India as locations for global R&D is likely to be a hazardous oversight.

The potential of China and India to serve as platforms for global innovation goes well beyond the availability of talent. These two countries face some of the world's toughest challenges as well as unique opportunities. Given the scarce supply and rising prices of raw materials and energy that is already evident, given also global warming, it will be utterly impossible for China and India to provide much better housing, much better food, much better transportation, much better education, and much better health care for their 2.4 billion citizens with today's technologies and business models.

Consider banking. About 70 percent of India's population lives in villages, and less than 5 percent of these have access to a bank branch. In short, in India today, about two-thirds of the population (nearly 700 million people) has no access to any bank of any type. It is hard to imagine that the solution will come in the form of brick-and-mortar branches of the sort that are common in much of the world today. Instead, the solution will almost certainly come in the form of extremely low-cost and ubiquitous mobile banking. Given current trends, it is a given that within three to four years, at least half of the adults in rural India will own a mobile phone and have access to mobile voice and data services. It is possible that India (and, similarly, China) will emerge as hotbeds of innovations in large-scale, secure, and extremely efficient mobile banking. Banks and telecommunications companies that see this opportunity and go after it may well end up leading the world in mobile banking solutions.

Procter & Gamble (P&G) provides another example of business model innovation with detergents in China. Not unlike Coca-Cola's syrup model (where the company supplies concentrate to local bottlers), P&G has started distributing high-value performance chemicals to local partners, which add the basic ingredients and packaging before distributing them to local retailers and consumers. The result is huge savings in transportation and energy costs, cheaper products for end consumers, and local jobs. Might this business model be transferable to other emerging markets, such as India, Southeast Asia, Latin America, and Africa? Absolutely. In fact, as consumers in developed economies start feeling the bite of higher raw material and energy costs, might some of the business models that originate from China and India have leading-edge relevance also for developed markets? As Alan Lafley, P&G's CEO notes, Asian firms are now “very innovative, especially with business models.”⁶

Transportation provides another example. According to Goldman Sachs, the number of cars on the roads in China and India is likely to rise from about 30 million to 40 million today to about 750 million by 2040.⁷ From the perspective of materials, energy, and emissions, growth at this scale is impossible if the technology in cars in 2040 is even remotely similar to that of today. There are only two possible solutions. One is that raw material and energy prices become prohibitively high and economic growth in these two countries comes to a screeching halt. The other (and the one that we bet on) is that China and India will become the hotbeds for breakthrough innovations in technologies, products, processes, and business models. Robert Lutz, head of product development for General Motors, is probably not off the mark when he notes that “cars powered by fuel cells (devices that use hydrogen to make electricity) are likely to take off in China before they do in the United States.”⁸

Realizing the Opportunities

Leveraging China and India as hubs for global advantage requires making decisions and taking actions along several fronts:

• Undertaking an activity-by-activity analysis of the value chain in order to decide which specific set of activities should be based in China, which in India, and which in other countries: for example, IBM's decision to use India as the global hub for IT services and China as the global hub for procurement.
• For each activity that will be based in China or India, deciding whether to set up the company's own operations or rely on external vendors: for example, Apple's decision to rely on subcontract manufacturers for the assembly of iPods in China.
• Ensuring that the China- and India-based global hubs will be able to develop the required world-class capabilities and, where important, cost structures: for example, Microsoft's efforts to cultivate relationships with leading-edge computer science departments in China to help ensure that it is able to recruit the best and brightest graduates for its research center in Beijing.
• Ensuring that as the company's value chain becomes more disaggregated and geographically dispersed, it will have the required visibility into and control over the activities of various players so that standards of quality, ethical behavior, and other practices are upheld throughout the value chain: for example, Wal-Mart's efforts to ensure that its several hundred suppliers in China and India meet the company's quality requirements, behave ethically, and pursue ever-more environment-friendly policies.
• Building the integrating mechanisms that will enable the hubs in China and India to be responsive to the needs of the company's global operations and enable the output of these hubs to be deployed to subsidiaries and customers in other countries: for example, the structural and budgetary links between GE's research centers in Bangalore and Shanghai and its business divisions.
• Managing the evolution of China- and India-based hubs as these operations become larger, more experienced, and more mature, on the one hand, and as China and India become more expensive, on the other: for example, upgrading the status of internal or external manufacturers within China and India from purely manufacturing operations to design and manufacturing operations.
• Managing the intracorporate friction that may accompany the shift of resources and power from the company's existing hubs in developed countries to the new hubs in China and India.

Disaggregated Value Chain Analysis

The starting point in figuring out how to leverage China and India for global advantage is to decide which activities in the value chain should be located within China or within India, or in both countries, and, in either case, at which locations within the country. Analytically, the way to arrive at the correct decisions is to undertake a disaggregated value chain analysis, which has three steps:⁹

Figure 4.1. Illustrative Example of the Value Chain in Biopharmaceuticals R&D

1. Draw an end-to-end map of all of the activities and subactivities in the value chain, starting with suppliers at one end to customers at the other. The more finely grained the map is, the better the subsequent analyses are likely to be. Figure 4.1 provides an illustrative example of a typical value chain for the drug development stage in the biopharmaceuticals industry.¹⁰
2. For each activity or subactivity in the value chain, figure out the optimal architecture for the activity based on the following key criteria: quality and market responsiveness, total cost, and level of associated risk. We use the term activity architecture to refer to the following two decisions: number of locations and charter of each location. The four ideal types of activity architecture are (1) global centralization in one location, (2) multiple but differentiated centers of excellence, (3) regional decentralization, and (4) country-level decentralization.
3. On an activity-by-activity basis, once you have clarity about the optimal architecture for the activity, make decisions about the optimal set of locations. With respect to China and India, the analysis should result in clear answers to the following questions:
   • Which specific activities will not be performed within China or within India? For these activities, the company's operations within China and India will need to rely on peer subsidiaries in other countries.
   • Which specific activities will be performed within China or India, or both, on a local-for-local basis?
   • Which specific activities will be performed within China or India, or both, on a local-for-regional (for example, Asia-Pacific or all emerging markets) basis?
   • Which specific activities will be performed within China or India, or both, on a local-for-global basis?

An important question regarding the global optimization of the value chain pertains to the extent of granularity in the analysis. At one extreme, granularity can be relatively coarse. An example would be a U.S.-headquartered toy manufacturer that centralizes all product development in the United States and all manufacturing in China. At the other extreme, the granularity can be much finer. A stark example of fine granularity is a U.S.-headquartered company that provides outsourced technical support services to its client companies. This company has created three operations centers: in Manila (Philippines), Bangalore (India), and Boca Raton (Florida). When a customer calls the U.S. toll-free number for technical support, the call is routed first to a level 1 analyst in Manila. If this analyst is unable to satisfy the customer, the call is rerouted to a more specialized but also more expensive level 2 analyst in Bangalore. If this also fails to solve the customer's problem, the call is then rerouted again to an even more highly skilled but very expensive level 3 analyst in Boca Raton. All of this multicountry routing takes place during the same call from the customer while maintaining a high degree of continuity in the conversation with him or her. A senior executive who is also a cofounder of the company explained that it would be impossible for his company to compete without this three-layered architecture. Costs for the skills that this company needs are the lowest in Manila and the highest in Florida. The goal is to maximize customer satisfaction at the lowest possible cost. Thus, it matters greatly as to what proportion of customer calls can be handled satisfactorily at level 1, what proportion gets transferred to level 2, and what proportion to level 3.

As an example of finely grained value chain disaggregation in the context of a physical goods business, consider the case of Li & Fung, a Hong Kong-based company. Li & Fung supplies over two thousand customers from a network of about ten thousand suppliers spread over forty countries. Fulfilling an apparel order from a U.S. customer can mean that the fabric is woven in China, the fasteners are sourced from South Korea, and the sewing is done in Guatemala.¹¹

There are pros and cons to both coarse and finely grained disaggregation. Coarse disaggregation means that the complexity and the cost of coordination would be low, and there would be less risk of misalignment across complementary but geographically dispersed activities. At the same time, coarse disaggregation may lead to a less-than-complete optimization of location choices for the performance of the different activities. In contrast, finely grained disaggregation can lead to a more complete optimization of location choices but can also significantly increase complexity, raise the costs of coordination, and heighten the risk of misalignment. Thus, the final decision regarding the extent of granularity in value chain disaggregation must rest on how advanced the company's capabilities are at cross-border integration of the physical as well as the information supply chain.

Figure 4.2. Impact of Integration Capabilities on Ability to Optimize Location Choices

In an era where communications as well as transportation technologies are advancing exponentially, we advocate that companies should overinvest in the development of multicountry integration capabilities (a topic we discuss later in this chapter). The stronger the integration capabilities are, the greater would be the extent to which the company can atomize the value chain and push harder in optimizing the choice of locations. Figure 4.2 depicts the relationship between integration capabilities and location optimization.

Give Primacy to Quality and Responsiveness. In deciding whether and how to use China and India (and, similarly, other locations) as platforms for global advantage, the first assessment screen must be whether the China- and India-based activities will meet or exceed your standards for quality and responsiveness. Clearly cost structure is important. However, cost minimization can never take precedence over quality and responsiveness. Quality of output from a research lab, a manufacturing base, or a call center depends not just on the capabilities of the talent available at the location, but also on the company's own actions to further develop these capabilities. It also depends on how the company organizes and manages these operations. Thus, in making location choices, the company must look at both: available capabilities at the various locations and what these capabilities could become after proactive actions to develop them further.

Ensuring that the China- or India-based operations are responsive to global needs also depends on several factors: training, cultivation of mutual understanding, communication links, supply chain links, and others. In the case of production and service centers, an increasingly effective approach to build responsiveness while maintaining a low-cost structure is to adopt a dual-shore policy: locate the global production and delivery center in India or China, but build a second production and delivery center physically proximate to the market. The proximate centers take care of unplanned rapid-turnaround demands from the customer, whereas the global centers take care of long lead time but planned customer needs. The dual-shore strategy originated with the leading India-based IT services companies, which derive over 90 percent of their revenues from outside India but locate over 90 percent of their staff in India. These companies faced intense pressures for customer responsiveness and invented the dual-shore mechanism as a strategic response.

The dual-shore approach is now being embraced also by companies that manufacture physical goods. A leading example is Bharat Forge, an auto components supplier headquartered in India. Bharat Forge supplies components to almost all of the major auto companies in India, Europe, the United States, and China. Through a combination of greenfield start-ups and acquisitions, the company owns eight plants: two in India, three in Germany, and one each in Sweden, Scotland, and the United States. It also has a joint venture in China with FAW, the largest auto company in that country. Bharat Forge's acquisition strategy has been driven explicitly by the goal of ensuring that it can supply all of a customer's needs from two plants: one based in India and one based close to the customer. This approach significantly reduces supply chain risks, makes the supply chain more responsive to customers' needs, and still enables the company to benefit heavily from a significantly lower cost structure in India.

Focus on Total Costs. In making decisions about whether to use India or China as global platforms for any particular activity, it is also important to look at total costs rather than just the cost differential between performing the operation in India or China versus elsewhere. While this may seem like an obvious statement, we have come across far too many companies that get carried away by superficial analysis and make wrong decisions that they regret later. At a macrolevel, total costs include the following five components:

• Cost of operations at the hub within China or India
• Cost of coordination between this hub and other entities within the global enterprise, as well as relevant external players such as customers, suppliers, and business partners
• Supply chain costs pertaining to both physical and information goods
• Costs of misalignment between complementary but geographically dispersed activities in the value chain
• Taxes imposed or incentives provided (or both) by various governments.

Every one of these costs is a function of not only location and scale but also how the company is organized and managed. The more advanced a company's capabilities at cross-border integration are, the more sharply it will be able to reduce the costs of coordination, supply chain logistics, and misalignment. Nonetheless, these costs will never be zero and must be taken into account in assessing the size of the benefits that might accrue from using China or India as global hubs for any particular activity.

As an example of total cost analysis, consider the case of one of the world's largest consulting and IT services companies with a major business process outsourcing (BPO) hub in India. In a 2007 interview, a senior executive of this company identified the following as the three major elements of costs in his business:

• Infrastructure, such as buildings, utilities, and travel
• Technology, such as fiber optics and satellite-based communication links between the operations in India and other client and intracorporate locations worldwide
• Employee benefits and salaries

The first two elements of cost, infrastructure and technology, accounted for about 55 percent of total cost. These costs were relatively stable over time, although technology costs were declining and infrastructure costs were going up marginally because of inflation but with a downward pressure due to increasing economies of scale. The third element of cost, employees, accounted for 45 percent of total cost and was increasing at a 15 percent annual rate. Yet as he noted, the average total compensation of a BPO employee was about $7,500 in India versus about $45,000 in the United States. Thus, if salaries in India grew at a 15 percent annual rate ($1,225 per year over the next twelve months) and those in the United States grew at a 5 percent annual rate (i.e., $2,250 per year over the next twelve months), the absolute cost gap between the United States and India would still increase rather than decline over time.

Minimize Risk. The third major factor in deciding whether and how to leverage China and India as platforms for global advantage pertains to the impact of the different options on various types of risk: political risks, currency risks, intellectual property risks, natural hazards, and others. Depending on the activity and type of risk to which the company may be exposed, following are some examples of how companies can minimize their vulnerability:

• In order to reduce their exposure to policy and currency risks, many companies with major production centers in China are adopting a “China + 1” or “China + 2” strategy. Thus, the manufacturing base in China remains the primary production hub, and the company builds additional production hubs in locations such as Southeast Asia and India.
• In order to reduce their exposure to supply chain risks, an increasing number of companies are pursuing a dual-shore strategy, whereby every customer is served from two locations: one based in China or India and one based close to the customer.
• In order to reduce their exposure to intellectual property risks, many companies disaggregate the overall technology or product development project into subprojects and disperse responsibility for different subprojects across different locations and even different countries. Thus, deliberate (or, involuntary) leakage of intellectual property from any single location would cause only minimal damage to the company.

These are illustrative examples of how a company can minimize its exposure to various types of risk while still being able to leverage China and India for global advantage.

Insource Versus Outsource?

In August 2007, Eli Lilly & Co., the U.S.-headquartered pharmaceutical giant, announced that it had signed a research contract with Chi-Med, a China-based contract research company. As part of this contract, Lilly would hand over to Chi-Med a portfolio of synthetic compounds that may contain within them the potential to develop drugs to treat cancer and inflammatory diseases. Chi-Med would be responsible for preclinical research. Lilly would provide continuing support to Chi-Med as well as pay an upfront fee plus at least $20 million for any compound that emerges from this partnership. After the initial discovery by Chi-Med, Lilly will take over the potential drug candidates for clinical trials and clearance through the U.S. Food and Drug Administration. Chi-Med would also get a share of any royalties or profits that Lilly may earn from a drug that becomes commercially successful.¹²

It is easy to understand why Lilly would choose China for drug discovery research. Qualified researchers for this type of work in China cost about one-quarter of their counterparts in the United States or Europe. Also, the supply of talent is large and easily scalable. However, instead of contracting the work to Chi-Med, Lilly could have chosen to set up its own drug discovery center in China. Yet it did not. Thus, the question arises: What logic should guide companies such as Lilly in deciding whether they should set up their own global hub in China or India, or both, or rely on an outsourcing arrangement with an independent organization?

Another example is Portal Player, the Silicon Valley-based company that supplied the brains for the original Apple iPod. Portal Player set up its own R&D center in Hyderabad, India. Engineers based in Silicon Valley and Hyderabad collaborated around the clock to develop the microprocessor, satisfy an extremely demanding customer, and in the process help make history. What logic should guide a company (such as Portal Player) in figuring out whether an in-house R&D center in India was the smarter approach, or whether they would have been better off subcontracting the research to an India-based company such as Wipro, which has a thriving business in contract R&D? Did Portal Player succeed because of insourcing in India, or despite it?

Our own analysis, backed up by the logic of transaction cost economics,¹³ leads us to put forward two major factors that should drive companies to decide whether to insource or outsource global hub operations in China or India: modular decomposability and economies of scale and scope.

Modular Decomposability. The more easily, cleanly, and without controversy a task can be decomposed into two or more parts, the more modular the subtasks become. A high degree of modularity permits two independent companies to take ownership of the separate tasks, write enforceable contracts that define the commercial relationship between them, and minimize the likelihood of wasteful friction. At the other extreme, a lack of modularity would make it almost impossible for two independent companies to agree on which is responsible for what. In such a case, what is required is joint ownership and, thus, an insourcing approach. Going back to the Lilly and Chi-Med example, note that the interface between preclinical research and clinical research is highly modular. Thus, outsourcing was a viable option for Lilly. In contrast, in the Portal Player example, the need for around-the-clock collaboration between the Silicon Valley and Hyderabad teams implied that a clean decomposition of the overall R&D project into two parts was infeasible. Thus, insourcing emerged as the only viable option.

Economies of Scale and Scope. In cases where the modular decomposability of the overall task is high, both insourcing and outsourcing are viable options. In such a context, economies of scale and scope should determine the choice between the two. In the Lilly case, Chi-Med has already built a large-scale preclinical research operation in China that performs contract research for several of the world's largest pharmaceutical companies. In contrast, while Lilly enjoys economies of scale in preclinical research in other locations, this is not the case for it in China. Thus, relying on a contractual relationship with Chi-Med enables Lilly to tap into the Chi-Med's greater scale and scope within China. This reasoning also leads us to predict that as Lilly expands the scale and scope of its drug discovery activities in China, it is more likely to set up an in-house research organization within the country.

The importance of scale and scope economies is illustrated also by a study from McKinsey & Co. and NASSCOM, an India-based industry association of software and service providers. According to this study, captive (that is, insourced) back office service operations are on average more expensive than outsourced operations, without any compensating advantages in terms of either better-quality work or lower staff turnover.¹⁴ This is so because independent companies that provide outsourced back office services to several clients are generally much larger and thus more likely to enjoy economies of scale and scope than a captive operation with only one customer, its parent company. Economies of scale and scope manifest themselves not only in the form of lower cost but also in the form of better-quality work because a larger organization is able to attract better talent, provide more attractive career options, and build core competencies around its activity domain.

Building Local Capabilities

For any activity in the value chain, a direct corollary of the decision to locate a global hub in China or India is that the capabilities at this hub must meet or beat all four criteria: quality, responsiveness, total cost, and risk. This statement applies to every type of activity: a global research center, a global production base, a global marketing hub, or a global call center. As depicted in Figure 4.3, building the needed capabilities at a global hub will always require an optimal blend of global processes, global talent, and global resources with local processes, local talent, and local resources.

Because of huge variations across industries, company strategies, and specific activities, there can be no universal answer to the question of what mix of global and local would be optimal for a particular global hub. The answer must be derived separately in each case. It is possible, however, to advance a set of universal guidelines that can be used to frame the analysis and discussions that lead to deriving the appropriate answer:

Figure 4.3. Building a Global Hub

1. The creation of a global hub (or the transformation of an existing local operation into a global hub) must be driven by a clear mandate from the top leadership, including the corporate CEO.
2. The executive directly responsible for building the capabilities and resources at the global hub should be a believer in the idea and should have internal conviction that the local talent to be recruited for the hub will be capable of serving the global needs of the enterprise as well as or better than any other locations.
3. The senior leadership team of the hub, including the executive directly responsible for building it, must be deeply and simultaneously embedded in both ecosystems: the local ecosystem within the host country as well as the global ecosystem of the parent enterprise.
4. In building the global hub, the goal should be to accumulate the needed resources and capabilities as rapidly as possible while ensuring that the hub internalizes global processes and global standards.
5. As the resources and capabilities are being built, there should be a clear and well-communicated expectation that the hub is expected not just to implement the existing best practices of the enterprise but also emerge as a source of new leading-edge ideas for the global enterprise.

The case of how Accenture, the consulting and technology services giant, developed its global delivery capabilities in India provides an excellent example of how to build a global hub that can help the enterprise leverage the talent, cost, and innovation potential of India (or, similarly, China) for global advantage. Accenture entered India in 1987, and for the next fourteen years, its operations there consisted largely of a small consulting base targeted at the domestic Indian market. The turning point came in 2001 when Karl-Heinz Floether, a German executive who headed Accenture's worldwide Financial Services business and was also a member of the board of directors, sent one of his senior executives, Keith Haviland, to India with a mandate to start building an offshore group there. Haviland, a British citizen, reported directly to Floether and had built Financial Services group's existing delivery center in London. In short, Haviland started with a mandate from the very top and knew how to build a global delivery center. As Floether recalled:

Haviland co-opted some of his experienced colleagues from the United Kingdom and United States, and the team relocated to India. Within a few months, they had recruited the first two hundred staff members. At that point, Floether visited Haviland and his team. He also visited the campuses of some of the major Indian competitors, as well as other companies, including GE, Philips, and British Airways. The visit convinced Floether that he had seriously underestimated the quality and scale of the talent base available in India, as well as the competitive threat from Indian companies. At this point, he asked Haviland to pull out all stops and hire as fast as possible.

At the corporate level, Floether was also successful in making the case that India-based global delivery was going to be crucial not just for the Financial Services group but for every other operating group within Accenture. Soon after, the delivery centers in India were transferred from Financial Services to the company's Systems Integration and Technology organization, a global capability-centric organization which served all operating groups.

In a discussion on the various decisions and actions that helped ensure a successful and aggressive ramp-up of the delivery capabilities in India, Keith Haviland recalled the following as having played a particularly important role:¹⁶

• There was a clear mandate from the very top: first Karl-Heinz Floether and, shortly after, the board of directors.
• The senior leadership viewed the establishment of a major global delivery center in India as a long-term strategic commitment rather than a short-term experiment. This enabled Haviland and his team to recruit senior managers (many of them expatriate Indians) from existing Accenture operations in other locations who would return to India and lead the charge. The long-term commitment also made it easier for Accenture to recruit seasoned executives from competitors within India.
• From the early days, Haviland and his colleagues insisted that notwithstanding explosive growth, the Indian operations must adhere to Accenture's high-performance delivery model: the company's standard ways of doing things, standard methods and practices, standard measurement approaches, and so forth. Insistence on standard Accenture processes eliminated the need for reinvention and made rapid expansion easier.
• The founding team paid particular attention to ensuring that there would be no compromise in the standards for recruitment. In the early days, key human resource managers from London played a direct role in interviewing candidates.
• With backing from senior leaders, Haviland and his team started channeling client work to India. Two early clients were J. P. Morgan Chase and DuPont. Capitalizing on the company's excellent relationship with both clients, the leadership in the United States persuaded executives in these two companies to start experimenting with service delivery out of India. In the first stage, a team of software developers was sent on a short-term assignment from Bangalore to Chicago to work on projects for these clients. In the second stage, this team moved back to Bangalore and continued delivering services from there. Over time, these proactive efforts to start channeling client work to India helped ensure that the hubs in India were building not just human capacity but also proven delivery capability.
• As part of its corporate processes, Accenture implemented programs to assess external and internal customer satisfaction. The delivery teams in India were assessed regularly on their ability to satisfy external as well as internal clients on a worldwide basis.
• The delivery centers in India were pushed aggressively to achieve global credentials such as the Capability Maturity Model (CMM) Level 5, CMMi Level 5, ISO27001, BS7799, SAS70 Type II, and others. These credentials not only ensured that the hubs were developing required world-class capabilities but also eliminated any concerns that external clients or other units within Accenture may have about relying on India-based delivery centers.

Over time, as the momentum has fed on itself, Accenture's head count in India has grown from fewer than three hundred in 2001 to over thirty-five thousand by 2007. India is now the largest Accenture country in the world, with some of the strongest technology and consulting capabilities in the global enterprise. Several of the company's senior executives (such as Sandeep Arora, Lead-Delivery Centers for Technology in India; P. G. Raghuraman, Lead-Delivery Centers for BPO in India; and Harsh Manglik, chairman and managing director, Accenture India) are now based in India. It is widely believed that over the coming years, the parent company will derive many of its global leaders from its India-based operations.

Another example from a relatively smaller company in a different industry illustrates that while the details will always be context specific, the guidelines regarding how to build the required capabilities at a global hub in India or China are quite universal. Pacific Trade International (PTI) is a $100 million revenue home decor products company, headquartered in the United States. Founded in 1994 by David Wang and Mei Xu, PTI serves a blue-chip roster of retailers, such as Target, J.C. Penney, and IKEA. Its operations are based in three locations: the United States (design, marketing and sales), China (design and manufacturing), and Vietnam (manufacturing). PTI's two primary business units are Chesapeake Bay Candles, which focuses on decorative candles, and Blissliving, which focuses on home furnishing products such as beddings and decorative pillows. We take a closer look here at Blissliving, the newer of the two business units.

Launched in 2005, Blissliving operates two design centers: one based at corporate headquarters in Maryland and the other in Hangzhou, China. The Hangzhou design center is co-located with the production facilities. This co-location plays a critical role in helping accelerate speed of product development, ensuring that production technologies keep up with design requirements, and permitting greater degree of customization. Here is the sequence of steps that Blissliving followed in building the needed design capabilities to support its rapid global expansion.

1. The company hired two young French designers, fresh out of design schools in France, who would move to Hangzhou for a one- to two-year assignment. It also hired five local designers in China. The charge to the French designers was to train their Chinese colleagues rather than to lead the design work.
2. Simultaneously, the company hired a product manager and two designers in the United States. All three came with prior experience working with Chinese manufacturers.
3. As depicted in Table 4.4 the company followed a deliberate and systematic approach to build the needed mindsets and design capabilities at the hub in Hangzhou. As this table indicates, in the first season, the U.S. team executed all of the key steps in the design process. However, with each subsequent season, the responsibilities became distributed between the U.S. and the Chinese teams, with the latter taking on an increasing share of the responsibility.
4. By the end of six seasons, the China-based team had developed the ability to lead and execute the entire work for a collection.

Blissliving now has two design teams, one based in the United States and the other in China, each able to execute a complete design project on a global basis. The expanded dual capacity has become a significant asset to the company in expanding the scope of offerings for various markets around the globe. Mei Xu, PTI's president, explained the philosophy that guided the buildup of design capabilities in China:

Table 4.4. Cultivation of Product Design and Development Capabilities at PTI's Design Hub in China

Ensuring Global Integration

In addition to building the necessary capabilities at a designated global hub, an equally important task is to ensure that these capabilities will be responsive to and serve the needs of other units and external clients worldwide. As depicted in Figure 4.4, effective and efficient integration requires a combination of hard and soft linkages.

The hard integrators serve as the foundation. They provide the enabling mechanisms for integration to occur. At the same time, how these mechanisms are used (and whether they are used or abused) depends on people. Thus, the soft integrators play an equally important but complementary role to the hard integrators.

Figure 4.4. Linking Mechanisms to Integrate a Global Hub with Other Units of the Global Enterprise

Information Technology Infrastructure. Given the rapidly growing power, increasing user friendliness, and declining cost of technology, we take it as given that it is better for companies to err on the side of over- rather than underinvestment in information and communications technologies. Companies such as IBM and Accenture, which can handle the IT needs of any global enterprise on an outsourced basis, are making access to a sophisticated IT infrastructure easier and less expensive. Without getting into the technical details of what would constitute sound decisions regarding IT infrastructure, we advocate the following:

• It is critical to have a common IT architecture (such as e-mail systems and databases) across the entire global network. No amount of sophistication in soft integrators can overcome the challenges that would be created by an inability to access real-time reliable data effortlessly.
• Companies should maximize the intracompany deployment and use of Web 2.0 technologies such as Wikipedia, Facebook, blogs, idea markets, and the like. These mechanisms permit large numbers of people to collaborate with each other easily, voluntarily, and inexpensively.
• Companies should constantly strive to increase the bandwidth of available communication links. A leading example today is Cisco's TelePresence, which permits near lifelike high-fidelity video communication between geographically dispersed individuals. The cost of high-bandwidth communication continues to decline dramatically.

Supply Chain Links. The goal of any well-designed and well-functioning supply chain system must be to ensure forward and backward visibility (into critical information such as demand forecasts, pending orders, available supplies, and production forecasts), responsiveness to fluctuations on both the demand and the supply sides, and, of course, efficiency on a total cost basis. As value chains become more finely disaggregated and companies rely on a larger number of production centers (as well as a larger number of external suppliers), the importance of forward and backward visibility rises. This is so because extended value chains increase the number of points where defects (such as lead in the paint used on a small part of a toy) can enter the production network and cause catastrophic failure later.

A senior executive at Nokia Corporation highlighted the importance of an effective and efficient supply chain management system in helping the company realize the competitive advantage being created in the company's R&D labs:

Formal Organizational Links. Formal organizational links include a variety of mechanisms. We look at three of the more important ones: reporting relationships, incentive systems, and budget allocation processes. In these areas, organizational decisions should be driven by a bias to induce integration of the hub with the rest of the global enterprise. A global hub that is poorly responsive to the needs of peer units in other countries becomes a deadweight rather than a source of competitive advantage.

With respect to reporting relationships, the central question pertains to whether the hub should report to the country manager locally within China or India or globally to the worldwide activity or business unit leader. Our perspective on this question is clear. While there will always be multidimensional matrix-type linkages, the primary reporting relationship of the global hub should be to the worldwide activity or business unit leader. A case in point is Microsoft's research operations in Beijing. True, the R&D center is located in China. However, as a global hub, its customer base includes Microsoft worldwide, not just Microsoft China. Thus, the correct decision about reporting relationships is to have the head of R&D in Beijing report directly to the head of worldwide R&D in Redmond, Washington, while maintaining a dotted-line relationship to other units in China, including the president of Microsoft China.

Another illustrative example is Accenture, which has multiple global hubs in India: a global hub for delivery of BPO services, a global hub for the delivery of technology services, and a global research center. The leaders of each of these global hubs report (as they should) directly and respectively to the worldwide leader for BPO services, the worldwide leader for technology services, and the worldwide head of research. Based in India, they have a close working relationship with each other and draw on support and guidance from the chairman of Accenture India. However, these local linkages represent lateral rather than hierarchical relationships.

As a third example, GE's John F. Welch Technology Centre in Bangalore belongs to and reports directly to the worldwide head of GE Research. It has close lateral linkages with other GE units in India. However, as a global R&D center, it gets its marching orders from the head of global research rather than the head of GE India.

With respect to incentive systems, key decision makers at the global hub must have a stake not just in managing internal activities at the hub but also in ensuring that these activities lead to commercial success by peer units in other countries. An effective way to build this responsiveness is to decompose the incentives for hub leaders into two parts: one part driven by functional excellence in how the hub is managed and the other driven by the financial results of the product lines and business units that depend on this hub.

An illustrative example is the Shanghai-based design hub for the Blissliving product line at PTI. An ideal incentive system for the designers in Shanghai would be as follows: an annual salary increase that is a function of labor market conditions in Shanghai as well as the designers' technical capabilities at their core task of design and an internal royalty-type bonus determined on the basis of the revenues generated by the products designed by them. Such a two-part incentive system is likely to ensure that the designers value both design excellence and market responsiveness.

Finally, the role of budget allocation processes in ensuring that a hub would be responsive to the needs of peer units in the rest of the global network is illustrated well by GE's technology center in Bangalore. This center gets half of its annual budget directly from corporate headquarters. These funds are allocated on the basis of beliefs regarding core technologies that would be of long-term and broad-based importance to most of the company's businesses. The other half of the budget comes from contracts signed between GE Research and the business divisions. This part of the budget automatically ensures that the technology center has to be responsive to the needs of the business divisions.

Social Networks. As we know from everyday experience, the effectiveness and efficiency of interaction between any two individuals are functions of not just technology links and formal organizational relationships, but also whether the people involved know each other as human beings and whether they like and trust each other. A considerable body of research has confirmed that social networks play a major role in fostering business relationships within the enterprise as well as across interfirm boundaries. Interpersonal familiarity and trust play a particularly important role in whether people are willing to share nonroutine information and their own judgments about the real story behind the hard numbers. While these observations are universally true across all cultures, the importance of social networks is even more important in the case of global hubs based in China and India. In both cultures, interpersonal relationships are known to play a significantly greater role than in Western societies. In short, the global enterprise must do its utmost to maximize the likelihood that the interaction between, say, a marketing manager sitting in Atlanta in the United States and a production head sitting in Hangzhou, China, will be between “John” and “Qiang” who know each other personally and can interact on a first-name basis rather than between “Mr. Smith” and “Mr. Wang” who see each other largely as occupants of formal organizational roles.

Social networks can be fostered through multiple channels. Some of the more common channels are periodic in-person visits to each other's locations (which should be designed to also include social interactions outside of the work environment), collaborations in cross-border business teams, joint participation in management development programs, and, in the case of high-potential employees, cross-country job rotations that foster professional development while also building social networks.

Managing Internal Politics

Notwithstanding economic rationality, the creation of global hubs (for R&D, production, back office services—you name it) is often fraught with internal struggles within the company for resources, jobs, decision-making power, and prestige.

A company's operations in China or India often start by producing for the local market or by offshoring low-end commodity tasks such as labor-intensive assembly or back office operations. In these early stages, the China- or India-based activities may rightly be viewed by people within the company (including senior management) as relatively peripheral to the company's core. However, a major transformation in mindsets, allocation of key resources, and distribution of internal power becomes essential if the China or India operations are to shift from the periphery to the core. This shift may happen in one or both of two ways: a major scale-up in the magnitude of the China- or India-based low-cost operations or a growing reliance on China and India for high-end knowledge-intensive activities such as research, product development, and business model innovation. Scale begets power. Knowledge too begets power.

Making operations in China or India core to the company's global competitive advantage need not necessarily be a zero-sum game. Whether it is depends on two factors: the company's growth rate and the speed with which it needs to build up China- or India-based capabilities. For a rapidly growing company such as Google, the buildup of R&D activities in China and India can easily and accurately be portrayed as a non-zero-sum enhancement of the company's capabilities. In contrast, for a relatively mature company such as IBM, it is much harder to avoid the perception that the buildup in, say, India will not substitute for current capabilities in, say, the United States or Europe. Speed of buildup also plays a role in that a slower buildup of China- or India-based capabilities may allow senior management to deal with the fallout in existing operations through normal attrition such as retirements and turnover. On the other hand, a rapid buildup may make layoffs in existing locations necessary.

Regardless of whether a buildup of China- or India-based capabilities will affect jobs in existing locations, it is clear that as these two countries become more central to a company's global competitiveness, a shift must occur in the structure of power over key resources and decisions. We have yet to come across a company where this shift in resources and decision-making power took place without direct and enthusiastic support from the company's CEO along with one or more of his or her colleagues on the corporate executive committee. If the CEO lacks the conviction that China and India are core to the company's future competitive advantage, the rest of the organization is bound to remain frozen in protecting the status quo. Even when the CEO has the conviction and the courage to act, he or she will often need to rely on a powerful senior colleague to lead the charge. The histories of how GE, Microsoft, and IBM built global competence centers in China or India, or both, illustrate the powerful role that enthusiastic support from the CEO can play in this process.

GE's John F. Welch Technology Centre (JFWTC) in Bangalore is now the largest of the company's four global R&D centers: Niskayuna, New York (staff size: 1,500), Shanghai (staff size: 1,500), Munich (staff size: 100), and JFWTC, Bangalore (staff size: 4,000). The story leading up to the establishment of JFWTC began in the mid-1990s when R. A. Mashelkar, then director-general of India's Council for Scientific and Industrial Research (CSIR), approached GE to inquire about any problems that the country's national laboratories could work on. GE's plastics business gave CSIR some of its tough technological problems. Within a year, the National Chemical Laboratory came back with a solution to one of the problems that GE Plastics had worked on for years. At that point, Gary Rogers, president of the plastics business, reported that given this talent, he would like to set up a plastics R&D center in India. Jack Welch, who knew the plastics business inside out, responded by arguing that if this reasoning was valid for plastics, it ought to be valid for the whole of GE.¹⁸ In 1999, the John F. Welch Technology Centre was born. Welch attended the opening ceremony and noted, “India is a developing country, but it is a developed country as regards its superb scientific infrastructure. It is for this reason that we wish to shift a part of GE's development effort to India.”¹⁹

Microsoft's research center in Beijing, Microsoft Research Asia (MSRA) is the largest research center outside Redmond in the company's global R&D network, which also includes research labs in San Francisco, Silicon Valley, Cambridge (United Kingdom), and Bangalore. The center opened in 1998. Barely six years later, in a 2004 article, MIT's Technology Review magazine hailed MSRA as one of the hottest computer labs in the world.²⁰ According to a detailed history of Microsoft's trajectory in China, the leading champion behind the setting up of the China research operations was Nathan Myhrvold, group vice president and chief technology officer, who reported directly to Bill Gates. As Myhrvold would later recall, “Bill was quite in favor of it. That was the wonderful thing about what I did at Microsoft—Bill was very supportive and understood the value of making long-term investments. And so that's what we set out to do.”²¹ In a company such as Microsoft, it is hard to imagine much opposition for an initiative whose champions included Bill Gates and Nathan Myhrvold.

Consider now the case of IBM, whose staff in India soared from fewer than ten thousand in 2004 to over seventy thousand in 2007 and is still adding over ten thousand people a year. Sam Palmisano, who took over as CEO in 2002, knew the services business thoroughly, having served as president of this business for several years under his predecessor, Lou Gerstner. Thus, Palmisano knew firsthand what was happening in the worldwide services business and the rapidly growing threat that Indian IT companies would pose to IBM and other incumbents. He did not need to be persuaded that IBM in India must become as large as or larger than the Indian IT companies, and he became the lead champion behind the transformation. As part of helping IBM's stakeholders think differently about the new reality, Palmisano has used every occasion that he can to promulgate the concept of an integrated global enterprise. As he noted in a 2006 speech at INSEAD, the French business school:

As these cases illustrate, enthusiastic and explicit support from the very top is critical if the company is to leverage China and India as platforms for global advantage. Without such support, these two countries are likely to be treated largely as an “n + 2” phenomenon, two additional markets to go after, rather than as transformational for the whole enterprise.

Conclusion

As we noted in the opening chapter of this book, China and India represent four stories: rapidly growing megamarkets, platforms for global cost reduction, platforms for global innovation, and springboards for the emergence of new fearsome competitors. In this chapter, we have analyzed the second and third of these four stories, looked at the factors that make China and India meaningful platforms for global advantage, and put forward a road map for what a global enterprise must do to capture these advantages. Which of the four stories will be more critical at any point in time is likely to vary across industries as well as over time. Thus, the wise approach for any company is to pursue a multitrack strategy for these two economies.

GE provides a case in point. It entered India in the 1980s with the goal of looking for market opportunities in businesses such as lighting, appliances, power generation, and medical systems. However, the company discovered that the market opportunities in India at that time were pitifully small. At the same time, it discovered huge opportunities to leverage India's IT and R&D talent for its global businesses. GE built an immense IT services operation in India as well as the Bangalore-based John F. Welch Technology Centre. The market opportunities in India have taken off, and GE is now aiming for $8 billion to $10 billion revenue from India by 2010. We believe that GE's multitrack mindset has universal relevance for other companies regardless of industry, country of origin, or company size.