Cash Flow

Does Microsoft Need China?

The tech giant is on the rebound, but its future may lie in how it decides to adapt its pricing model to the developing world.
Tom LeanderAugust 10, 2004

If you looked in the mirror one morning and discovered you were Bill Gates, how would you regard China? With a hard-won sense of balance, tempered by infinite patience, perhaps. But more likely with the frustration of a champion player cut from the title match.

Microsoft, the US$36 billion software company co-founded by Gates, famously got off on the wrong foot in the PRC, and the troubles afflicting its presence still abound. It owns the desktop market there, yet earns little money because 97 percent of its software is illegally copied. Prominently among Asian governments, China’s state planners support Linux, the open-source—read non-proprietary and low-cost—operating system that competes with Microsoft’s Windows. Every time Microsoft pressures the government to crack down on the pirates, the government makes a move to support the rival system.

And yet Microsoft is pumping US$750 million in aid into China over three years—and that’s on top of the approximately US$1 billion it spends there annually to run its business. The money is devoted to helping develop an infrastructure for a software industry via joint ventures and academic research and training.

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

Microsoft is helping China as much—or more—than any US company. Still, among the Western companies seeking a fortune there, Microsoft seems to face the greatest obstacles. If you were a prominent Microsofter, you might ask yourself, “Does this company need China?”

Darn Shame’

“No,” says John Connors, CFO of Microsoft, when asked this same question. Connors, who took over the top global finance post in 2000, is speaking to business school students and local executives in Singapore.

At 45 years old and more than six feet tall, he has a youthful face and a voice that still has a hint of a drawl from the Montana rodeo town where he grew up. “No,” he says. “We can still be successful without large-scale China growth. Our financial model, our business plan doesn’t assume any big China breakthrough. There’s nothing on the multi-year that [shows] it breaks through.” He adds: “But it would be a darn shame if it didn’t.”

On the face of it, this is true. Nobody needs China less than Microsoft. It’s a company with US$36 billion in revenue, total assets of US$90 billion, and 57,000 employees. It has an astounding US$56 billion in cash (Connors plans an announcement late in July about the cash, and analysts expect a share buyback). True, it’s no longer a high-tech dynamo whose stock price is growing at 40 percent per annum. In an admission of slowing growth, CEO Steve Ballmer recently announced US$1 billion in cuts. Yet analysts still project an 11 percent rise in stock price this year. This company, if anything, is resilient, and midway through its fiscal 2004, analysts are still in a glowing mood.

David Hilal, an analyst for Friedman Billings Ramsey, a US investment bank, projects revenues to grow 8 to 12 percent and earnings to increase 11 to 15 percent per year on average over the next five years. “I’m bullish on Microsoft and its earnings power,” says Hilal. “They’ve shown their ability to drive top-line growth and made progress in containing costs.”

But in the long run, China could pose dangers to Microsoft. If Linux flourishes there, it could spawn formidable low-cost rivals to the American company. “The real value of open source to a country like China,” says Kevin MacIsaac, an analyst with the MetaGroup in Sydney, “is developing a public infrastructure for a software industry. It’s a reasonable and cost-efficient way for China to compete globally.”

Others in Asia see the potential. Japan and South Korea joined China in April on a project to jointly develop a new operating system based on Linux as an alternative to Microsoft’s Windows. Thailand and Malaysia have instigated programs to offer low-cost PCs to citizens with Linux operating systems (see “The Butterfly Effect on Global Pricing?”). They’re being helped along by Microsoft competitors such as Sun Microsystems, which has signed a deal with the Chinese government to supply its Linux desktop operating system and office program to as many as a million PCs there. Future electronics products shipped from China—such as mobile phones and DVD players—could be developed free from dependence on the Windows operating system.

For now, however, Microsoft is in a sweet-enough spot. Demand for its products in the US and other mature markets is strong enough to offset piracy problems in China and elsewhere in the developing world. A restructuring in 2003 has kick-started a cost containment program spearheaded by Connors. The company switched from reporting based on geography to seven discrete product groups, appointing seven CFOs to run them as separate businesses. The company’s anti-trust battles have also eased. Most recently, a US appeals court approved Microsoft’s 2001 antitrust settlement with the Justice Department, effectively closing a six-year battle. While an antitrust battle is ongoing in Europe, the company put aside its long-running fracas with Sun Microsystems earlier this year, with a US$2 billion settlement.

Interviewed in May, Connors presents the company’s global future with measured optimism. “We have a good presence in a very small part of the business market today on the desktop,” he says, “and a very small share of the overall budgets, as well as a small share of the consumer market in terms of things that are going to be digitized. And we have a very small share of the market in terms of small- to mid-sized business apps.” He adds: “We think that overall markets will grow at some rate better than world GDP. If we can grow the business better than world GDP, and the business we’re in is growing faster than world GDP—that’s a good place to be.”

It all depends, he says, on what happens to price. “As long as volumes grow and prices remain relatively firm or even soften a bit,” says Connors, “we can have a great business from the revenue perspective, and we can also have a good business from a cash generation or a value generation perspective.”

All Eyes on Price

Put another way, Microsoft is relying on current pricing and a goodly portion of the world’s tech growth to sustain its 31 percent net profit margins. But an increasing portion of global tech growth will come from Asia’s burgeoning economies. And it’s precisely in Asia—with China in the lead—that pressure to alter the uniform pricing structure for its software is the strongest in the world.

“I would like to use your software, but how can I invest in it when I now have cheaper options available?” asked a CFO of a healthcare chain based in Singapore at an executive breakfast where Connors was speaking that morning. The CFO added: “What can I tell my shareholders?”

Connors responded that the total cost of ownership of Microsoft Windows and Office products—which account for 80 percent of its revenue—is in fact less than that of cheaper, open-source software, because Microsoft can offer the entire weight of the ‘eco-system’ that supports its products. This eco-system can be described as the support, customization, integration services, and software that evolve around the Windows product. Connors cited studies that have endorsed this view from Forrester Research and Merrill Lynch (see “Two Views on Total Cost of Ownership”).

But while Connors’ arguments hold true in the Western world, they don’t stand up so well in emerging markets. That’s because Microsoft has a one-size-fits-all pricing policy for the world. And while it offers discounts in various sectors—such as to governments and universities—it remains inflexible on the cost of its software. Critics say that the total cost of ownership argument does not apply in markets like China, where labor costs are cheap and the real cost of Microsoft products is extraordinarily high. It’s partly for this reason that Microsoft has such serious piracy issues in Asia.

Research gives an idea just how much Microsoft products cost in the region. A group based at the University of Maastricht in Holland called Free/Libre/Open Source Software (FLOSS) released a study last year that showed the cost to businesses of a basic ‘toolkit’ of Windows XP and Office XP in 173 countries using purchasing power parity rather than market exchange rates. For data on the countries, the FLOSS study drew upon World Bank figures published in 2001 using a figure of US$560 for the Windows and Office toolset if it was purchased in the US, based on contemporary pricing from The FLOSS study said the same package would cost a Chinese business US$21,678, equivalent to more than seven months of the nation’s per capita GDP. In India the cost came to US$42,725, or 14 months of per capita GDP. In Cambodia, the price tag reached US$71,184, or 24 months of per capita GDP. Says Rishab Ayer Ghosh, program leader for FLOSS: “In developing countries, even after software price discounts, the price tag for proprietary software is enormous in purchasing power terms.” Ghosh argues that lower labor costs and higher licensing fees tilt the debate about the total cost of ownership between open-source and proprietary software in favor of open source in most developing countries.

In this context, Linux presents Asian governments with a low-cost opportunity to develop a competitive market for software, causing downward pressure on price. Security-minded governments like China’s see a benefit in developing a home-grown software industry independent of the proprietary code of the largest American software company. Moreover, a lively Linux market, if it developed in China, would provide a solution to the nation’s software piracy problem. Illegal users could naturally gravitate to the lower-cost Linux software rather than buy Windows at seven months per capita GDP.

All of this amounts to tremendous bargaining power against Microsoft.

Is It the Same?

Microsoft bashing is so common worldwide—and so much of a knee-jerk reaction—that it’s almost iconoclastic to look at how the familiar devil isn’t so familiar anymore. For one, on the finance side, Microsoft has turned itself into a more agile company, and in a remarkably short amount of time. Two years ago, Steve Ballmer, Microsoft’s CEO, suggested to John Connors that the company should restructure along the lines of General Electric into separate businesses reported on separate P&Ls. (The Wall Street Journal reported that Ballmer had his revelation after reading former GE CEO Jack Welch’s memoirs.)

Connors has enacted the change with remarkable swiftness for a company so large. Last summer, he appointed seven CFOs to the seven P&s, which are business solutions, server and tools, mobile and embedded devices, home and entertainment, MSN, client, and information worker. The seven finance chiefs were given the power over setting targets, performance measurement, and budgeting. Analysts said the new structure represented Microsoft’s maturity from a tech wunderkind to a giant industrial company.

Says Hilal at Friedman Billings Ramsey: “The move to seven P&Ls and seven CFOs was aimed at treating each product group as a separate company—and they’ve seen benefits.’ This year, MSN joined the client, information worker, and server and tools groups as Microsoft profit engines, Hilal says. “The three businesses that are losing money are losing a lot less of it,” he says.

To Connors, the move represented a shift in which finance was given an equal footing in a company which had always been the province of techies and ‘the field’. “We now have these P&Ls constructed over and above the field,” says Connors. “In the past, we had our geographical people who managed their own P&Ls—and we didn’t have them managing R&D. Now we have P&L leaders in charge of R&D, sales, services, and marketing.”

Still to be done, says Connors, is implementing “this in a way that our field isn’t frustrated.” He elaborates: “Say Australia has a great program in mind for a promotion that isn’t the same as France and Germany; we need to make sure that they’re not stopped from doing that.” He talks of a need to “make a big investment in public sector sales and relationship work. Making sure we have that kind of rapid decision-making movement as we would have before we established the seven PGs (product groups).”

Learning to Adapt

The Microsoft conundrum – of an admired giant seeking to find a footing in the developing world—is perhaps most succinctly defined by a Microsofter himself. “I have a company that economically is being driven by the mature market,” says John MacLellan, the company’s finance director for the Asia Pacific and greater China regions based in Singapore, “and I have to understand how to drive it by the emerging markets.”

MacLellan sees the seven P&L structure as a boost to his efforts in China. “We have to realize that we have 97 percent piracy and whatever we do we’re not going to get economic returns – not in the short term,” he says. “The seven P&L structure allows us to communicate with our shareholders more clearly about the bets we’re making,” he says. In other words, the strategy for X-Box—Microsoft’s game console—in China is not joined at the hip to the strategies for the other units, and the goals of investments within a given group can be argued on their own merits.

MacLellan is a Scot amid his second stint at Microsoft. He says a prime reason for his return was the lure of Asia and making Microsoft work here. “There’s no higher priority than getting Asia right,” he says. Despite his background in finance, he has the zeal of an expert salesman, though he hardly talks the party line. “We’re taking this phenomenal 800 pound concept,” he says, “and we’re trying to fit it in a region where this 800 pound concept may not fit.” He adds: “Of course, if you’ve been here for five minutes you see that it doesn’t. And we’re having to adapt.”

That adjustment involves getting China in a different way, understanding the dynamics between the central and local governments, how businesses struggle, what China needs. “We in the West seem to regard the direction of China with a McCarthyite view,” he says, “as if there’s a unified, ominous intent behind official decisions.” MacLellan is an engaging conversationalist, a tall man whose size makes a stronger impression because he’s an animated, passionate speaker. “But this isn’t the case,” he continues. “We have to have the patience to let them figure some of this stuff out, and to understand the bravery of some of their decisions.”

He adds, “It’s a phenomenal engine that these guys are trying to manage, and trying to understand. They’re looking for the levers.”

The same could be said about Microsoft in its encounter in China. Things started badly when Microsoft opened an office in Taiwan in 1989 and began shipping its Windows products into the PRC. A Taiwanese programmer had inserted patriotic statements into the software, which were revealed later and offended the Chinese government. Then, as the magnitude of the piracy problem developed, Microsoft launched a series of lawsuits, which alienated public opinion. Bill Gates also told Fortune magazine in 1998 that eventually Chinese users would become ‘addicted’ to computers and become paying customers. The phrase is used by Pogo Linux, an open-source software company based, like Microsoft, in Redmond, Washington, in its Chinese marketing materials. “Chinese executives don’t miss the parallel with the Opium Wars and China’s victimization at the hands of the colonial powers,” says Tim Lee, president of Pogo Linux.

But then Microsoft reversed its tracks. Its US$750 million investment in the PRC over three years (beginning in June 2002) is an extraordinary bid to help China build a market for software development—and help put Microsoft in better standing with the government. The deal allows China’s state planners a say in where Microsoft donates millions of dollars worth of technology training. It calls for Microsoft to buy hardware for its X-Box consoles from Chinese partners, and the government gets to pick the partners. The money is being used to set up four joint-research labs at Chinese universities and to pay teachers at universities elsewhere in China.

All this, of course, is more than altruism. In the absence of a developed ‘eco-system’ for its products in China, Microsoft is determined to create one. “The question is how we can help build a software and IT ecosystem,” says Connors. “China is a long way behind in the eco-system. If you look at India, it’s very Silicon Valleyish. China doesn’t have that advantage in IT or software.”

The attraction for a government like China’s, he says, is a multiplier effect that surrounds Windows products. This multiplier is the company’s estimate of the business and earnings generated by all the independent software development, support, customization, and integration that surrounds Windows. Microsoft pegs that multiplier at between 7 and 8 times the amount of the cost of licensing. “That’s tens of thousands of people,” says Connors, “and to the extent that they can continue to make money, they’ll do so.” He adds: “If we do a great job, the multiplier effect is good.” In Microsoft’s argument, the mass purchase of Windows provides a boost to any economy.

Says MacLellan: “There’s a danger that people see us as taking too much of the ecosystem, when in fact we’re trying to do the opposite. I’m going to use the M-word and use it very carefully. I want the ecosystem to monopolize on Microsoft. I want as much of the eco-system built on my platform, so that more people say, ‘I’ve got to be on the Microsoft platform to get access to all of these people.’”sion-making movement as we would have before we established the seven PGs (product groups).”

Whose Addiction?

These words do in fact sound as if Microsoft is trying to create a dependency on its products, a natural enough goal for any business within legal bounds. But critics challenge the assumption of exclusivity. Eco-systems are necessary in software, but they do not, prima facie, rely on Microsoft to develop them.

It may be closer to the truth that Microsoft is in a race to establish the dominant eco-system. Certainly, Linux supporters are many years behind. They may be praying for some form of Miracle Gro to juice the Linux eco-system into full flower before Microsoft’s overwhelms it. Time, some say, is on Microsoft’s side and it could still win in China. Dion Wiggins, vice president and research director with Gartner Group, a technology advisory, points out that Microsoft’s situation in the PRC is not as bad as it looks. In a recent note, “China, Intellectual Property, and the Big Picture,” he argues that China is following the pattern of Western nations in its treatment of intellectual property protection and is slowly evolving a legal infrastructure that will allow foreign companies to press their cases in Chinese courts. Politically, it’s in China’s interest to control piracy now that it is a World Trade Organization member.

Any clampdown would lead to an advantage to Microsoft. “They own the desktop in China, in terms of sheer market saturation,” says Wiggins. “The problem is the conversion from non-paying to paying customers.” The company is not without solutions. Among the most radical but intriguing would be a one-time amnesty for pirate Windows and Office users, allowing current offenders a one-shot deal at a license.

The idea here would be to convert these millions into instant, paying customers—with the dividend of appearing to endorse China’s IT development. Another route, of course, would be a concession on price steep enough to attract pirated software users to make the switch. This would fly against the current thinking in Redmond, and is unlikely given the company’s current stance. But there have been precedents in other industries. Apple offered downloads of music from the internet as an alternative to illegal downloads on MP3 players, but did so within a price range that music lovers could afford. It found that many customers were willing to go legal.

And besides, says Wiggins, a cheaper, localized Chinese version of Windows wouldn’t be a threat to Microsoft worldwide. “The export market demand for Chinese language versions has to be limited,” he says. “What real use does a Chinese, Thai, or Lao version of Windows have in Germany?”

Microsoft, to be sure, is not unaware of the pricing issue in developing markets. The programs launched by Thailand and Malaysia to sell low cost computers to citizens with Linux operating systems have prompted it to offer a Windows XP Starter Edition – a stripped down version of Windows – to customers in those two countries at a lower cost. The company followed up with an announcement that it would offer the starter version throughout Asia. Analysts like Wiggins doubt whether the starter version will be cheap enough to combat the rampant piracy.

Still, it’s a start. Eventually, Microsoft will have to face the pricing issue in the developing world head-on. For now, it has the luxury of time on its side. No one knows how long this will last.

“They’ve been growing and doing fine without China,” says Hilal. He adds: “It’s a problem if they can’t make money in China, but what the magnitude is, is difficult to say.”

Does Microsoft need China? Maybe not. At least not now. Does it want to win there? You bet. And its future could well depend on securing that victory.

Two Views on Total Cost of Ownership

Microsoft has long argued that licensing fees from software represent only 5 percent of the total costs of an IT system, and that broad range of costs must be factored into any realistic assessment. Lately, a body of research has been generated that supports Microsoft’s contention that Windows is the better bargain Ð at least in the developed world, and for now.

There are various methodologies for viewing the total budgeted and unbudgeted costs over the lifespan of an IT system. These are divided into direct costs and indirect costs, with the direct, or budgeted, costs breaking down into hardware and software, IT operations, and administration. Indirect, or unbudgeted costs, are categorized as user operations and computer downtime.

Several studies this year – among them by tech group IDC and Forrester Research, both based in the US – have shown that total costs of operations favor Windows over Linux. The IDC study, for example, showed Windows had a cost advantage over five years of between 11 and 12 percent for four out of five typical server workloads. Linux proved to be cheaper in one category only: simple web serving. Those who favor Linux point out that some of the studies, such as the Forrester research, was funded by Microsoft, although no one has questioned the independence of Forrester’s approach. A portion of Microsoft’s website,, is given over to the findings of these reports.

The critiques of Linux mostly boil down to issues emerging from Linux’s comparatively early stage of evolution. While Linux may be attractive on several fronts, not enough apps built on the operating system exist to support some business tasks as well as its rival.

The rebuttal from the Linux crowd is that it’s a matter of time before the eco-system—everything from a critical mass of software applications to specialists who can support Linux compatibility at affordable prices—develops. IBM, Oracle, and Hewlett-Packard have all embraced Linux and will help build toward this critical mass.

For now, it appears that Microsoft is ahead in the TCO debate only in developed markets. Rishab Ayer Ghosh, program leader of Free/Libre/Open Source Software, a think tank at the University of Maastricht in Holland, argues that exorbitantly high real costs of licensed software, coupled with lower costs of labor in the developing world, tips the TCO argument in favor of Linux.

The Butterfly Effect on Global Pricing?

A scheme to bring computers to the masses in Thailand is leading to region-wide pressure on Microsoft to reconsider its pricing model, in the view of a some analysts.

In May 2003, Prime Minister Thaksin Shinawatra announced a program to provide 1 million low-cost personal computers to Thai consumers through the Thai Ministry of Information, Communications, and Technology. The idea was to provide low-income earners with the means to buy a PC with internet access and a loan package to finance the purchase. Microsoft originally rejected the Thaksin government’s invitation to participate by including a reduced-price version of its software in the program. Microsoft did not offer a Windows XP or Office XP Standard, Internet Explorer, or Windows Media Player in a Thai language version at the time of the program’s inception, and this was instrumental in it’s decision to stay out.

The Thai government opted for Thai-language versions of Red Hat Linux and Sun Microsystems StarOffice in lieu of Windows. Though many called it ambitious, the program quickly took off, and Microsoft’s marketshare began immediately to drop. By summer 2003, Microsoft had performed an about-face, designed a Thai-language version of its products, and quickly bought into the program— offering its software at a radical discount of US$40 (the discount applied to Windows XP Home Edition and Office XP Standard Edition). Gartner Group has reported that by the time the program rolled into its second and third phases, the majority of applicants were choosing Windows.

Dion Wiggins, research director for the Gartner Group in Hong Kong, believes that the Thai precedent could have far-reaching implications. In a report last summer, he likened the incident to the butterfly effect, renowned in chaos theory. The term refers to how tiny events balloon into macro changes over time. Will this incident remain the flutter of a wing—or lead to a hurricane? Malaysia’s government, seeing the success of the neighbor’s program, launched a similar campaign this year and Microsoft participated. Eventually, it may be tough to contain so many butterflies.

The Price in Asia

The chart below shows license fees relative to GDP/capita for a basic toolset of Windows and Office XP*, and the effective US$ equivalent for the toolset across Asia, expressed via purchasing power parity rather than market exchange rates. The sources used are the World Bank World Development Indicators Database, 2001. Piracy data from the Business Software Alliance, a trade group, is shown to demonstrate the co-relation between piracy and higher real-costs for the software.

Country Annual GDP
Per Capita
Number of
PC’s (000)
Piracy of Microsoft Windows Cost for basic Windows toolset Cost GDP per capita months
United States US $35277 178326 25% US $560 0.19
Bangladesh 350 254 N/A 56401 19.19
Bhutan 644 5 N/A 30668 10.43
Cambodia 278 18 N/A 71184 24.21
China 911 24222 92 21678 7.37
Hong Kong 24074 2600 53 821 0.28
India 462 6031 70 42725 14.53
Indonesia 695 2298 88 28412 9.66
Japan 32601 44311 37 606 0.21
Korea 8917 12142 48 2215 0.75
Lao PDR 326 16 N/A 60525 20.62
Macao 14089 79 N/A 1402 0.48
Malaysia 3699 3000 70 5341 1.82
Mongolia 433 35 N/A 45598 15.51
Nepal 236 83 N/A 83770 28.50
Pakistan 415 585 83 47630 16.20
Papua N.G. 563 298 N/A 35071 11.93
Philippines 912 1702 63 21658 7.37
Singapore 20733 2100 51 953 0.32
Sri Lanka 849 175 N/A 23257 7.91
Thailand 1874 1698 77 10540 3.59
Vietnam 411 933 94 48011 16.33
Regional Aggregates:
Asia 2128 102229 N/A 9282 3.16
European Union 20863 116997 N/A 947 0.32
Latin America 4335 18703 N/A 4577 1.55
Africa 652 7636 N/A 30297 10.31
Middle East 2679 9708 N/A 7375 2.51

Source: Researched and compiled by Free/Libre/Open Source Software (FLOSS), University of Maastricht, Holland
*US price taken from