Mary Meeker on China

WSJ writes about Morgan Stanley’s Mary Meeker, dubbed “Queen of the Net”, and her growing interest in China:

After the U.S. Internet bubble burst in 2000, Ms. Meeker fell off the analyst radar screen and was assailed as having overhyped Internet stocks. But now, in the thick “China Internet Report” released Wednesday, Ms. Meeker is back, pleading that “investors still underestimate the impact the Internet will have in changing business process and consumer behavior on a global basis — and we believe that China is emerging as a market that helps prove this point.”

“The overall market opportunity from the Internet in China will prove to be quite significant over time,” she maintains. Moreover, she’s turned to China to prove her long-term theories about how the Internet can change business — especially through economies of scale, provided here nicely by China’s 1.3 billion potential consumers.

“One of the elegant things about the Internet is that you can get a little bit of money from a lot of people. That can create a pretty compelling business model,” Ms. Meeker says. Most Chinese people are poor but, Ms. Meeker argues, not too poor for the Internet. She points to the early success of Chinese online gaming companies, such as Shanda Interactive Entertainment, the owner of which became one of China’s richest men by charging cents for subscription access to games played by thousands at a time.

Ms. Meeker argues that simple math on a Chinese scale points to profit, even for companies following some of the same paths as U.S. dot-coms, such as online advertising and e-commerce. “If Internet user growth in China continues to grow at 30%, and usage grows higher as broadband becomes more pervasive, it’s fair to say the revenue growth from Internet companies should at least be commensurate,” she says.

Not everyone agrees. Ten-year veteran Chinese technology analyst Duncan Clark took issue with her assumptions that business models that have sustained U.S. dot-coms could work in China. “E-commerce can be a great window, but can you collect?” asks Mr. Clark, managing director of BDA China. “Now people are using sites just to say, ‘Let’s meet at this street to pay each other off.’

The 217-page report is accessible from here.

Another story in the WSJ discusses how online gaming is propelling growth of the Chinese Internet companies:

Highflying Chinese Web portals such as and tanked with the dot-com crash, but launched a spectacular comeback last year on the strength of selling text-messaging services. Those services, called SMS, for short messaging service, allow customers to send news, messages and pictures to cellphones through the portals. But now SMS may be playing second fiddle to sword fighting. This year, the prices of several Chinese Internet stocks have surged on the back of new revenue from selling online games.

Many of China’s online denizens aren’t slacking office workers but kids who, because of Beijing’s one-child policy, are frequently looking to entertain themselves at Internet cafes or home. Some 53% of Chinese Internet users are under age 24; in the U.S., only 29% of users are 18 to 29.

Online games, in which players subscribe to an Internet service that simultaneously delivers a game to thousands, are a natural because they don’t require users to have an expensive console. Last year, 15 million Chinese subscribers spent $160 million to play against real-life friends in virtual games, according to IDC analyst Jun-Fwu Chin. In 2008, he expects 54 million subscribers, who will spend $823 million. Further, online game companies have already proven they can make money by selling subscriptions for pennies. On April 2, Shanda — owned by one of China’s richest men, Tianqiao Chen — filed to have an initial public offering on Nasdaq, as Shanda Interactive Entertainment, that it hopes will raise $200 million. The company reported net income of nearly $28 million for 2003.

India will come on the radar as the broadband access infrastructure gets built, and is complemented by low-cost access terminals, affordable software and locally relevant content.

Linux on Desktops

The Economist writes:

Within the past month, some of the world’s most powerful technology firms have pledged considerable support for Linux on the desktop. Hewlett-Packard (HP), which runs neck-and-neck with Dell as the largest seller of PCs in the world, said it will begin shipping some machines that run on Novell’s flavour of the free operating system, called SuSE Linux. And Sun Microsystems, an arch-rival of Microsoft, announcedeven as it was preparing to bury the hatchet with Microsoft officiallythat it had persuaded Wal-Mart, the world’s biggest retailer, to sell cheap PCs using Linux and Sun’s StarOffice suite of application programs, instead of the ubiquitous Microsoft Office. This follows a deal that Sun struck last autumn with several Chinese ministries to ship up to 1m PCs with Sun’s Linux package to China this year, rising to tens of millions in future years, according to Jonathan Schwartz, Sun’s chief operating officer.

Is this the beginning of the end for Microsoft’s virtual monopoly on the desktop? Certainly not right away, says Al Gillen, an analyst at IDC, a technology consultancy. Today, almost 94% of all PCs in the world run on Windows, while slightly more than 3%mostly in creative industries and universitiesuse Apple’s Macintosh system. Fewer than 3% use Linux. By the end of the decade, Linux’s share could grow to 7-10%, reckons Mr Gillen, displacing Macs as the main alternative.

More specifically, two windows, so to speak, of opportunity appear to be opening. One is that the next version of Windows, called Longhorn, has been delayed to 2006 at the earliest, in part by Microsoft’s realisation that it has to tighten up security a lot more. So, for the next two years, companies and home users thinking about updating their operating system might be reluctant to buy the current version, Windows XP, knowing it will soon be overtaken. Hence they may consider alternatives more seriously. If Linux can establish a good reputation during this period, it might look even more attractive once Longhorn, which will be expensive and is likely to require new hardware, is released. Linux, after all, can be very cheap: $100 per user if bought as part of Sun’s package, for instance. It can even be downloaded for nothing from the internet.

The other opportunity lies in poor countries. Unlike rich countries, which have a huge installed base of Windows computers and billions of documents in Microsoft’s fiddly file formats, most users in Asia and South America are starting with a clean slate. For a country such as China, says Sun’s Mr Schwartz, the attraction of open-source software is obvious: it is cheaper, so it will reduce the incentive to get pirated software (most copies of Windows in China are fake) and thus help China improve its relations with the World Trade Organisation. Better still, it allows China to avoid being locked into a single vendorand an American one at that.

On the other hand, despite improvements Linux faces real obstacles. It can still be a nightmare for home users to install and, unless bought as part of a commercial package such as Sun’s, it does not come with a help-desk. Worse, there are still too few applications. Fewer than 1% of all computer games, for instance, work on Linux. Software to manage personal finances or organise digital photos is also missing. In theory these programs could all be written but, without a huge increase in users, code-writers will not bother.

The future is uncertain, and Linux still might yet represent another kind of threat to [Microsoft]. No standard operating system has yet emerged for mobile handsets, robots, watches, televisions, printers, car gadgets and other such devices. Microsoft, naturally, wants to extend Windows’ dominance to these as well. It is here, rather than the desktop, that Linux could be a real threat to the mighty company’s ambitions.

Broadband in India

The Times of India has a commentary by Pradip Baijal, who is the chairman of TRAI (Telecom Regulatory Authority of India), on what needs to be done to accelerate the process of broadbanding India:

The main reason behind our poor penetration is the abominably high subscription rates. In Korea , charge per 100 Kbps is a quarter US dollar per month. In India , the consumer pays 60 times more for the same speed. Korean GDP per capita is $10,000 against India ‘s $50.

Thus, on an affordability basis, the Indian consumer pays 1,200 times more. We must ensure that the charges are drastically reduced if they are not, the problem of low connectivity will never be overcome. Until last year, telecom tariffs in India exceeded those in China by a factor of 1.5, even though the GDP there was double that of India .

Before we plan for higher broadband penetration, we must analyse the present hindrances to growth. Of course, the current rates for consumers are very high. Growth cannot obviously happen unless these rates are brought down to nearer the Korean levels. There are other constraints as well. Broadband cannot be viably delivered by anyone other than the incumbent since the entire last mile connectivity is with him. Today, the local loop is not available to other operators. DTH, VSAT and wireless can deliver the last mile connection, but the costs of both DTH and VSAT are kept artificially high and wireless data networks are yet to spread. Since “within city” networks are solely with the incumbents, there is no competition despite the fact that there are many bandwidth suppliers.

National Internet Exchange of India was set up to ensure that local Internet traffic does not have to go through US, etc. However, the costs are very high in this network for smaller operators and inter-operator problems are not yet sorted out. Laying new last mile connection, always a problem, has been rendered impossible due to problems of right of way. Taxes and duties for broadband equipment are very high and there are few incentives for faster growth.

Business World had a detailed story recently:

After over Rs 51,000 crore of investment, the dozen or so companies involved don’t have much to show. All they have is about 100,000 subscribers who think broadband means just faster Internet access. In some of the world’s largest broadband markets – Korea, Japan, and to some extent the US – it is household penetration that set the broadband market on fire and made it profitable. In these countries, demand for entertainment in the form of either cable TV or gaming created a demand for fatter pipes at home, and therefore broadband. In Korea and Japan, telecom operators fulfilled this demand. In the US, cable operators did it.

Yet in India, hardly any push has happened to make telecom operators offer cable TV, or vice versa. Both have made some half-hearted forays, promising a lot but delivering just net access. “Our broadband penetration is just two per 10,000 people (0.02%), and we say we are a knowledge-based society,” exclaims Pradip Baijal, chairman of Trai and now the country’s broadcast regulator.

Baijal is communications minister Arun Shourie’s man in the field, and kick-starting broadband penetration is one of his priorities. He thinks that broadband can play a vital role in prodding GDP growth. The availability of the net, telephony and video on demand at low prices means that distance education, tele-medicine, tele-working, and research become easier. It means transaction costs for businesses go down, and overall gains in productivity.

two small paragraphs in two unheard of reports are broadband’s new hope. One is in a Confederation of Indian Industry (CII) discussion paper on the India Broadband Economy, and another is in a Telecom Regulatory Authority of India (TRAI) consultation paper. Both papers recommend that state-owned Bharat Sanchar Nigam (BSNL) and Mahanagar Telephone Nigam (MTNL) open up their last mile access to other companies.

Will it work? Maybe, just maybe, unbundling could trigger a fresh spurt of investment by telecom companies into the last mile. That, in turn, could push cable companies to offer more than just cable TV, regulation permitting. That, in turn, could push prices down, and so on. Whatever happens, there will be lots of confusion over the way it will work. There will be fights over who will invest and who will get a larger share of revenues. There will be endless debates on the technologies best suited to do this – DSL, cable or wireless – and more regulation may be needed in the initial phases. But for India’s frustrated broadband dreams, any hope, however slender, is welcome.

Bandwidth in India is expensive and still in the world of narrowband for the most part. Broadband has to become an integral part of India’s digital infrastructure.

SAP’s Challenges writes:

Among the threats to the German company are software interoperability standards called Web services, the entry of Microsoft into SAP’s market and growing interest in alternative software licensing models from upstarts like In addition, some securities analysts say SAP’s core market–major corporate software overhauls–has peaked, causing several of them to question the company’s growth potential.

Microsoft is hitting SAP in one of its weaker spots–the largely untapped market for business applications software geared toward small and midsize companies. A new software rental model also appeals to this segment.

Additionally, Web services enables large companies to stitch together otherwise incompatible applications, lessening the appeal of a monolithic system like SAP. Adding to this trend is the availability of low-wage software programmers in Asia and elsewhere, ready to build programs to specification.

While these challenges are daunting, no one is ready to write off SAP. With more than 21,000 customers–among them the global giants of commerce–SAP is practically guaranteed a steady stream of lucrative maintenance revenue for years to come.

SAP is still at the top of its game after one of the harshest periods in the computer industry’s history. According to AMR Research, SAP rakes in more than a third of the $20 billion or so that companies spend each year on enterprise resource planning applications–software designed to automate corporate tasks such as billing, payroll and customer service.

Contextual Leadership

HBS Working Knowledge writes: “Harvard Business School’s Leadership Initiative is attempting to answer these and other questions about the nature of leadership through its formation and study of the Great American Business Leaders database.” Excerpts from an interview with Tony Mayo, executive director of the Initiative:

When we examined our database of great business leaders by the decades in which they served their companies, the role of contextual intelligence became an increasingly compelling proposition.

. By analyzing our pool of great business leaders by decade, three distinct leadership patterns or archetypes emerged. We called these leadership archetypes Mold-Makers, Mold-Breakers, and Mold-Takers.

Makers essentially created or enhanced businesses that took advantage of the coalescing context of their times. They built businesses that thrived in a specific contextual framework and modified their operations and leadership styles as the contextual factors evolved. Some admittedly were in the right place at the right time, but it was often more than luck that made the difference between being merely successful and being wildly successful. In contrast, Breakers were business leaders who succeeded in breaking through the contextual mold of their times. They sought not to be constrained by the present conditions of their market; they were able to envision a future landscape for success. Finally, Takers were business leaders who saw value in industries and/or businesses that others thought were no longer viable. Takers took advantage of industry consolidations and often breathed new life into companies; extending business value well beyond what others thought was possible.

There is a strong tendency to search for a candidate who has a specific track record of success, but board members need to understand the context in which specific CEO candidates were successful. It is all too easy to ignore both the past contextual framework of success and the present one. Are they aligned? Does success in one context predict success in a new one? Would Sam Walton be as successful today? What about Michael Eisnerhas his time run out? Has he kept pace with the changing contextual framework? Despite the overwhelming evidence to the contrary, boards tend to favor the “proven” talent, but often fail to ask “proven in what context?”

TECH TALK: As India Develops: Distribution Hubs (Part 6)

The next excerpt from the RISC paper by Atanu Dey and Vinod Khosla looks at the economics:

The objective of a RISC is to provide as complete set of services as possible sustainably. Sustainability essentially imposes the condition that user fees (demand) have to cover the cost of the services delivered. Given the low average income in rural India, per capita demand for many of the services can be expected to be low. However, rural populations are heterogeneous in their demographic characteristics. A small but significant section of the rural population would be able and willing to make use of the services. Besides, there is a significant dynamic income effect.

The flow of incomes is straightforward in the model. Users pay for the services they need. The services are assumed to provide a net benefit to the users. The service providers in turn use infrastructural services and pay for them. The provision of services that are relevant to rural economic development, and do so efficiently, is the basic objective of a RISC.

The per capita demand for any single specific service in rural India has to be very low in keeping with the low per capita incomes. The RISC model derives its sustainability by aggregating the demand for a variety of services over a sufficiently large population. Thus the total amount of services required is a function of the aggregate income of the population rather than the per capita income.

If we assume an average daily per capita income of $1, then the annual income of the 100,000 target population of a single RISC is about $36 million. Assuming that the presence of the RISC actually increases productivity and economic efficiency so that the economic output of the population goes up by 10%, the increase in economic output will be roughly $3.6 million. Assigning half of this increased output as increased income to the population leaves nearly $2 million per year to pay for the services available at the RISC. The annual gross revenues per RISC will be around $2 million and the aggregate revenues for about 6,000 RISCs is approximately $12 billion per year.

The figures above only take into account the direct effect of RISC on the economy. This is only a partial analysis as there will obviously be a multiplier effect on the economy. Taking a conservative multiplier of about 2.5, the total effect of 6,000 RISCs will be about $30 billion in the initial years alone. However, realistically it may be that initially only a few states in the Indian union will be the early adopters. Assume that only about 2,000 RISCs are implemented in the first phase. Therefore the direct effect on the economy will be about $4 billion and the total effect on the GDP will be $10 billion.

Given annual gross revenues of $2 million per RISC, and assuming that profits account for 15% of gross, then given an interest rate below 15%, the investment per RISC can be up to $2 million. Given that land is relatively cheap in rural areas, that does amount to a lot of basic infrastructure in terms of investment in power, telecommunications, the physical plant and so on.
The most critical question regarding the implementation of RISC is about the source of the investment required for the infrastructure layer. The costs are significant. Assuming the capital costs of each RISC to be $1 million, then 5000 centers, enough to entirely saturate rural India, would cost $5 billion. While it is a large number in absolute terms, it is only about a percent of the annual GDP of India. Much of this investment is already being made, and we contend, somewhat inefficiently.

The central concept of RISC is that it is a market-based approach that requires the participation of the government (at national, state, and local levels), the private sector, NGOs and multinational lending institutions. Each of these have different (albeit overlapping) interests in broad based rural economic growth and therefore have an incentive to participate in the implementation.

The government has an incentive to expand social and economic opportunities in rural India. To that end, it already invests significant amounts in rural development. At the central level, the government already spends nearly 3% of GDP for various subsidies. The solution we present operationalizes the efficient deployment of the resources that are already being spent. In fact only a partial redirection of the current spending would in our view have a significant impact. Further it would attract additional investment from non-governmental sources if criticality is achieved.

By focusing the investment of various actors around these RISC centers, it is possible to gain the economies of scale, scope, and agglomeration. This agglomeration will also serve to mitigate risk in the investment.

Next Week: As India Develops (continued)

Continue reading