Tuesday, August 29, 2006

Is the World Flat?

I believe that India can become a strong economic power only through driving, controlling and dominating technology. To be able to do that, India has to be in the driving seat, be able to make decisions on direction of technology and be in a vantage position to foresee and deliver technology to the markets of the world. We need to do what Western Europe has done during Industrial Revolution, and what the USA and Japan have done in the Electronic, Automobile and Manufacturing sector in 20th Century. We have a unique opportunity here. Opportunity knocks once. Shall we embark on this adventurous ride or shall we miss it once again to be mere spectators and followers?

My previous article on ‘Why Infosys cannot become a product-making company’ was just a small part of the bigger problem-statement India has to deal with. India has to create its own technology ecosystem. To be in the driving seat one has to deliver the technology, goods and associated services from here in India.

Why I stress on such an ecosystem to be here in India?

Take examples of companies like Intel, Cisco, Siemens, Boeing, Toyota, and Tata. An ecosystem exists around each of them. The ecosystem consists of many other smaller or bigger companies that depend on each other. The larger ecosystem consists of academic and research institutes feeding these companies with new technologies and experts, a huge customer base in government deals or consumer markets, etc. Boeing in order to build planes supports and promotes ancillary industries- steel and aluminum, heavy electrical, communication equipment, chemical, tools, and so many others. Boeing by itself may be able to employ only 150,000 employees, but this industry in turn is helping many other industries.

Right now, some of these technology and product making companies use the services of many Indian companies- dealing with each of them independently to get the work done. The following picture illustrates this. Each of these Indian companies completes the project to enable this technology driver to make a product or develop technology. What an Indian company ends up is getting revenue for this project and also the experience of technology. Unless this experience is used to promote the same technology or product here in India, it is completely useless. This experience of working on a technology is extremely valuable and needs to be tapped into to generate maximum benefit. That can be done by creating similar technology and product making companies in the ecosystem which can generate the maximum value out of this experience.

Who is driving the technology in this picture? Who knows where and what will happen in three years from now. How many of the senior management at TCS and Satyam will be able to say with confidence what they will be working on and what they will be delivering three years from now? If indeed they know this, did they make the decisions or was the decision handed over to them from their client?

Thomas L. Friedman in his famous book titled The World is Flat has described the changes that are taking place in our world owing to fast communication and travel, bringing together diverse people of different countries (like India and China) to work together driving economic activity.

However, Friedman looks at it from an American (and West) perspective. His motivations are driven by what is good for America and the West. In some respects, his views seem to reflect that what is good for America is good for rest of the world too. Outsourcing and off-shoring are ‘in’ things. By stating this, he is preparing Americans and the western world to wake up to this new world order, embrace it, faster than what they hoped, and to take advantage of this new world order to perpetuate their economic dominance. In this new world order, countries like India and China may not be dominated like in colonial times, but will be ancillary partners supporting the plans of the western world helping them perpetuate their economic dominance.

Yes, the world is flat- it is flat for the western world, which drives the technology and the markets. For the followers and servicing organs like India and China, the world is still round. It is still complex, the boundaries still remain. Travel is still difficult and a good customer service is still far away. The cultural, political, religious issues continue to dominate the economic issues. May be, for a person living in Fremont (California), Bangalore is as near as Spokane (Washington), but for an Indian living in Tumkur, Bangalore is as far as Spokane. For most Indians the world order has not changed. Except for introduction of plastic (through sachets) and cellular phones to his village, the new globalization has not brought an American closer to him, the way it brought an Indian closer to an American. When an American picks up the phone to report a problem on his DELL laptop, he may be speaking to a considerate and American-accented Indian sitting in Bangalore, but when an Indian picks up the phone to report a problem with his water system, he is talking to a rude and apathetic operator. Not much has changed for this Indian the way it has changed for an American.

While it is extremely important for Americans and Western economic powers to wake up to this new flat-world order, recognize it and understand it so as to exploit and benefit from it, it is easy for Indians and Chinese to get misled by this interpretation into believing that the equations that hold good for an technology driver are also valid for him. That is not the case. While servicing these technology drivers in what is seen by them as a flat-world he is benefiting from it in a very narrow scheme of things. By not reaching the same level of being in the driving position he is losing out on the bigger scheme of things. This new flat-world order continues to keep the western powers at the helm using some select few developing countries as supporting and ancillary partners. It does not however bestow upon these select few developing countries the role of a leader. An Indian and Chinese should not get carried away by the interpretation of this new flat-world order into believing that he is an equal partner. He should not quickly conclude that he has an equal role to play in his respective segment (services- software or manufacturing). An inherent hierarchy of control still exists.

Many Indians, living in India and US, seem to like this new interpretation that indeed they are all on the level-playing field. This illusion seems to make them happy and they seem to join the bandwagon to champion this notion further. This is convenient for them because without having to be the leaders, they get the label as equal partners. This notion (of flat-world) is supported by everyone in the current industry. Software services organizations in India are happy, the new entrepreneurs promoting similar software services industry are happy. The VCs who promote this industry in India are happy. According to them- “In this flat world, India has a role to play. China has a role to play. We are all playing our roles. What is wrong with that?”

Take one consequence of this flat-world- Outsourcing of BPO has positive and negative fallout. While it is enabling many a youth to make a quick buck who otherwise would never been able to dream of making such money, it is also creating a youth which is confused, having the lowest sense of professional ethics, and seems to bring in a social structure which the Indian family is not used to. The question remains- is this a sustainable business? Outsourcing is a good stepping stone to something more sustainable, tenable and long-lasting. But what is that? Are we moving in that direction?

The world is flat, but the world is flatter for some than others.

I do not agree that the World is Flat for everyone. I think it is flat for those countries and companies who are in the driving seat. For all those countries and companies who are followers, the world is still round. Having multiple locations (in different countries) for a services company like Infosys, Wipro or TCS is not going to make this world flat for them. If indeed they believe this, they are fooling themselves. Putting offshore units in China or Vietnam is not going to take them anywhere. Boundaries exist in these parts of world which still stand.

R&D organizations should be made of closely knit groups. All my experiences working on multiple R&D locations (in different countries) were major disasters. The projects kept on getting delayed running into years and many of them seriously failed. Only those locations which kept their entire R&D team for a technology or product closely knit were able to deliver with success.

A software services company like TCS might be bagging many contracts from different companies in various domains. How many other companies in India get benefited from these contracts? How many other companies and industries does TCS spawn, help, aid or promote? How many companies depend on contracts that Infosys brings home? The outsourcing business is a like a pipe between the technology provider and the services company. It flows between the driver and the follower but no ecosystem get created. The technology provider at the driving seat decides who he will partner with for outsourcing. Such collaborations do not create an ecosystem in India. Once the cost-arbitrage is compromised, this technology provider may choose another partner for outsourcing and that could very well be in another country. For all we know he may even choose to stop this outsourcing business altogether.

Outsourcing across continents and countries is a gap filling exercise. New technology, like internet, has given us tools to try out different experiments and that’s what we are doing. The problems associated with long-distance and different locations will soon dominate and this whole exercise will be reexamined.

Champions of Indian services industry may look at examples like IBM, Accenture, EDS, etc, as role models. How many companies can you create out of India which can emulate them? Ten? Twenty? But that is not enough for us to become an economic powerhouse. NASSCOM predicts Indian IT-ITES will be a mere $70B industry by 2009. Is that good enough? We need to be thinking of making it a $500B industry. We need to be thinking of making technology producers and product makers in India who can earn large monies. We need a Nokia, a Boeing, a Toyota, a Oracle, a Apple, and a Dell. To conveniently skip this process terming it risky is a lackadaisical approach. Satisfying ourselves saying that we are services only will not allow us to scale as a nation. We need more technology drivers.

By employing 1 million engineers we may be able to take it to $40B. Using the present model of services, by employing 5 million engineers we may be able to take it to $200B in the next ten (or twenty) years. Now, Compare with this. Google with 6,000 employees is a $6B company. Nokia with 40,000 employees is a $40B company. With 1 million engineers we should be aiming at making $1 trillion. That’s when this country will be a developed nation.


Shashi Tharoor: Imperfections in Friedman's Flat World

Infosys: Think Flat Blog

Friday, August 18, 2006

Why CDMA will die!

I believe that by 2011 (in another five years) the market share of CDMA family in the world will be less than 2%. There are many analysts out there who do predict that CDMA market (CDMA, CDMA2000, etc) will shrink but their estimates are quite generous. Currently, the market share of CDMA family in the world is around 18% but is dwindling gradually (while GSM family currently holds 81% of the market). I believe that this decreasing rate will soon pick up pace to shrink quite drastically in an exponential fashion leading to its ultimate demise.

Some of the trends in various markets in the last few months:

SK Telecom and KT Freetel (in S. Korea), who collectively represent about 10 per cent of CDMA subscribers in the world are not likely to migrate to CDMA2000 1x Release A- according to Morgan Stanley report

Reliance and Shyam Telecom in India, China Unicom in China and AT&T in US are switching to GSM from CDMA.

VIVO (Brazil), the fourth largest CDMA carrier in the world has openly discussed the potential of migrating to the GSM roadmap- according to Morgan Stanley report

According to Telstra (of Australia), ‘CDMA is on the way out and other carriers will follow Telstra's lead because we have the second biggest network in the world.’ H3G, Australia’s only other CDMA network operator, is preparing to move away from CDMA as well.

Alegro PCS (of Ecuador) plans to migrate to GSM technology from the CDMA 1x platform currently in use.

According to a Credit-Suisse Report, "The share of CDMA subscribers in India would drop to 7 per cent by 2010, while that of GSM would grow from 75 per cent at present to 93 per cent.''

Isn’t CDMA a better technology compared to TDMA-based-GSM?

During the time (early 1990s) when 2G networks (GSM, TDMA, CDMA) technologies were being deployed, there has been much hype about various technologies and its advantages so much so that it became difficult to understand what is artificial and what is real. While it was clear to many theorists that CDMA as a technology was efficient in spectrum usage it was hard to determine its benefits in real world. Qualcomm solved many practical problems of CDMA technology to make it a practical technology. CDMA family as championed by Qualcomm is proprietary in nature and its business is based on royalty models. There is a limit to the progress made in a proprietary technology. With fewer players, minds and resources tackling the problems, the solutions are not easy to come compared to a collaborative technology. Though the evolutions of GSM (UMTS, etc) are based on the same technology (wideband CDMA) they are collaborative in nature and not proprietary. There have instances where a better technology has lost the race to an inferior technology.

When does a better technology lose race to an inferior technology?

There is more to business than a better technology. And sometimes a better technology may lose market while an inferior technology may grab huge market share. Some of the reasons are the following:

  1. Timing

A better technology coming later may not find itself a place in the market because the older technology though inferior may have the market penetration which the newer technology may find difficult to replace. Example, Signal Processing based on Wavelets though proved superior could not replace the Signal Processing based on Fourier Signals for Digital TV because the technological breakthroughs required for making it possible came later (after adoption of standards).

  1. Ecosystem

A technology has to have a good ecosystem which will keep innovating, drive the prices and costs lower, and increase the penetration into market. A good ecosystem will have many players working on the same technology- many startups innovating and decreasing the costs, big companies investing huge sums of money for development and deployment resulting in volume-based cost reductions, many customers and partners embracing this technology leading to market penetration.

  1. Volume

Volume once it has reached certain levels automatically drives the costs down. Once the R&D money is recovered the volume-based business usually lowers the costs making it attractive to the customers fueling market penetration. A newer technology though superior may not be able shake such volume-based businesses.

  1. Cost of Development and Deployment

The cost of development of a certain technology (during R&D and production) and the cost of deployment (cost in making it available to the customer) are important components in keeping a technology alive. Two competing technologies with different cost of developments will have different growth paths. A technology which offers superior user experience and features but with higher costs may find itself unattractive compared to a competing technology with inferior features but with affordable pricing.

Every technology faces competition- some survive and some die out. If a technology keeps improving itself, updating itself and is flexible in its business model then there are greater chances for its survival. However, a new technology though lacking all the above may still replace the older one if it introduces a completely new way of life which wasn’t available before (like the wheel or air travel). But be assured that it will face competition from a newer technology later in time and will be tested on the above parameters once again.

CDMA family is going to die out

Qualcomm’s CDMA is going to die out because it lacks in (2) Ecosystem and (4) Cost of Development. It hasn’t been able to develop the ecosystem because of its proprietary nature. Its business model on getting royalties on various forms of its technology hasn’t been able to spur many new companies and startups that could have benefited in promoting the technology to create this ecosystem. Innovations could not happen at the same pace (that of GSM family) because it was closed to new companies. Research and Development efforts at various big companies were bogged down by patent infringements cases and other royalty obligations. Qualcomm went into war with many companies including its partners and customers.

The cost of development is also high because of huge barriers (to entry) making it less attractive for entrants including newer and smaller companies. The cost of development, owing to less innovations, lack of ecosystem, and royalties, is higher compared to developing GSM family equipment- both infrastructure and handsets.

I don’t think the ‘royalty issue’ raised by Reliance (India) attributing it to higher cost of handsets is the only issue in making this strategic decision (of slowly moving to GSM family). It’s not the handsets alone but the complete infrastructure (and upgrades) which is expensive in CDMA family. Qualcomm gave lot of concessions to operators during deployment phases but expected to recover them through amortization that included sharing of profits from the operators. Once the operators started to feel the burden they became uneasy.

On the other hand, GSM family benefited from incremental innovations and marginal differences in each stage of development of the products resulting in a compounded effect of lowering the costs at all levels. The cost of developing infrastructure equipment and handsets is reducing drastically in the GSM family. Over a period of time, this difference in development costs, which did not seem obvious because of many incentives during development and deployment, became apparent as the operator started feeling the pain. Moreover, the user experience was lacking in CDMA family compared to GSM family. The operators who saw these trends started looking for alternatives (in this case, opting for GSM family). However, these operators may not switch right away because their skin is already in the game. The bigger operators can make the switch sooner and faster though incurring some losses, but the smaller ones will find it difficult to switch right away. The opportune time for this switch is when an operator has to overhaul its network (like moving from 2G to 3G, etc).

What is the future of Qualcomm?

I think that Qualcomm will transform itself into a 3G/WiMAX company contributing to advancement of these technologies producing chipsets for handsets, etc. It is already making strides in HSPA technologies and its recent acquisition of Flarion will also give it an added advantage in contributing towards OFDM based technologies. In my opinion, Qualcomm should just join others in promoting WiMAX instead of trying to create another market backing Flarion-based proprietary OFDM technology. There is another hurdle that Qualcomm has to cross to be able to make an impact on WiMAX (if it indeed joins WiMAX). It has to put efforts towards an effective and successful integration of Flarion, its people and technology, into the existing organization. Many great companies have floundered with acquisitions, and a cult like following that Qualcomm promotes may find it extra difficult in integrating Flarion into its fold to make any effective contribution.


CDMA family includes CDMA, CDMAOne, CDMA2000 and other evolutions like EVDO, 1xRTT, 3xRTT, etc, and are based on Qualcomm’s CDMA technology. This is currently deployed by operators like Verizon and Sprint in US and Reliance and Tata Indicom in India.

GSM family includes GSM, GPRS, EDGE, UMTS, HSDPA/HSUPA and is deployed by operators like Cingular and T-Mobile in US and Airtel and Hutch in India.

Some Links: [1], [2], [3]