Wednesday, April 18, 2007

Why do we have so many jobs in Bangalore?

I want to provide a different perspective to this topic. I touched upon it earlier at ‘How many people report into you?’ Here, I want to linger on a little longer.

Software-services companies and MNC offshore units inherently introduce inefficiencies that are supposedly alleviated by increasing the headcount, which while benefiting the group does the damage of killing the individual. What do I mean by this? Let me explain.

A division or group in a software-services company makes more money when it has more number of people in it. Therefore, the division head, the project manager, and the project leader will all collude to ensure that headcount keeps increasing. This attitude is set in early on and at every stage. Those who resist will either have to conform eventually (so that they can succeed) or they will be weeded out. Over a period of time, what you have is set of successful individuals who have mastered the art of inflating the number of resources to do a project. Say, a manager X says he needs 8 people to do a certain job, while another manager Y says he needs 12 people to the same job. Invariably, the manager Y is selected for the job. Give this process few years- what you have is a set of managers who are all set to outdo the other in inflating the numbers. Only those who inflate the numbers with panache and flair succeed.

Now, what happens at the vendor who is outsourcing to these software-services companies? There is a competition within the vendor company too, between different outsourcing managers, to outsource more. Say, there are two outsourcing managers (OMs) outsourcing to two different companies. There are two factors that come into play here. First, the OM who outsources more work will prove that he has saved more for the company, setting a trend to outsource even more. This outsourcing comes at a price though. There is a homegrown antipathy towards such outsourcing since each job outsourced means one less job for the local guy. But the senior management looks at it from a cost-saving perspective and goes ahead to reward the guy who saves more (who would eventually become the senior management). More outsourcing means more headcount at the software-services company. Second, the OM becomes the champion of the software-services company. He develops a giver-relationship with the company he outsources to. Being the outsourcing manager, he is treated like a king at the software-services company. He is the messiah, the giver and friend, all combined. The software-services company people look upon him to increase their share of revenues from the vendor. He in turn likes the attention he receives and takes it upon himself to do better to earn their respects and obeisance. Also, their success is his success. He becomes their savior and protector, and in turn helps himself. There is a bond that is established between the OM and software-services manager which in turn helps in increasing the headcount at the software-services company.

As a result, you will end up with divisions of 400 people to service a vendor in US/Europe when the same work can actually be done by 100.

Now, let me also look at MNC offshore units in Bangalore. Though they are an integral unit of the parent R&D, these offshore units are usually given the step-daughter treatment. The best work is not given to these offshore units – only the non-critical portions and other support, maintenance portions are assigned to these units. (Only few MNC have actually started to treat their Indian counterparts as mainstream R&D centers). The decision power is not shared either – the heads at these offshore units are mostly paper tigers, with great titles but little influence. Those who work at MNCs share a sense of frustration for not being able to get the best work and for not being able to influence. That leaves most of them to inherit and borrow the practices of other software-services companies of India where they worked before- the usual routine of political maneuvering and one-upmanship. This one-upmanship usually involves having more people ‘under you’. The more people report into you the more powerful you are.

Added to this, the influence or the contribution of a group, including its IP, is measured by the headcount rather than the actual value it produces. So, in effect, a group of 3 producing a superior IP is valued lower than a group of 20 producing an inferior IP. Therefore, you are not rewarding those who produce a superior IP with less number of people, but instead, you introduce a mechanism to reward the mediocrity. The head of offshore unit, bereft of any key decision making power on the overall strategy and business of the MNC, cannot show progress either in terms of revenues nor profits. In absence of these parameters, he resorts to showing progress in increase of headcount. Hence, the tendency to learn and perpetuate the art of inflating the numbers!

[Please note that the inflation of numbers in these offshore MNCs is not as high as in software-services companies since there are more checks and balances.]

I refer to such inflating-the-headcount practices as ‘mediocrity-breeding’ mechanisms. These practices do not award the star performers. They do not allow good performers to feel proud of their achievements. Star performers get disenchanted. They tend to award those who deliberately and smartly inflate the headcount requirements which actually increase inefficiencies. These practices tend to become virtues in both software services companies and MNC offshore units. Such environment does not take care of two fundamental things an organization should do- challenge the employee, and take care of the employee. Higher salary turns out to be the only incentive, which can always be used by a competing company to lure any engineer. This also leads to unprecedented levels of attrition. Solution to every lagging project or bad quality product or inefficient program seem to be addition of more people, as if, adding more people is suddenly going to alleviate the situation. Most often, such addition compounds the problem it is trying to solve. However, a suggestion to increase the headcount is more acceptable than a realistic toning down of the size.

Over a period of time, you have Bangalore, with hundreds of thousands of jobs, which benefits the group as a whole but has already killed the individual spirit.

Saturday, April 14, 2007

On Tax Holidays

Most of our IT software-services companies enjoy tax holidays. While this has been a good incentive that encouraged many such companies to proliferate in India, I think its time to take relook at this. Why should certain big companies, which have gone public, have a brand name, and are making colossal profits, be enjoying this tax holiday? Yes, I understand why it came into existence in the first place. We were not on a level-playing field, we needed support, encouragement to compete with global giants. But some of these tier-1 software-services companies have achieved the status of being able to compete with these global giants. Do they still need these tax holidays?

I propose that once a certain IT company reaches a revenue-mark, say $1B, combined with certain net profit-mark, say 20%, it should start paying taxes. Come to think of it, the government could use these funds in innovative ways. It could use some of these funds to create more technology oriented and product-oriented companies, like giving loans at cheaper rates, or investing in those companies on part-loan, part-equity model, creating zones and lab setups for such tech startups, or reimbursing money for patenting, or decreaing employee taxes to tech-startup in the first three years before it start making revenues, etc.

Hmm… we need bold, aggressive and innovative methods to go that next stage!

Infosys Results

Infosys has announced its results. It made approximately $3.2B last year with 72,000 employees. That amounts to approximately $44,000 per employee. That’s an improvement from the previous year where it made approximately $40,000 per employee. It plans to add approximately 25,000 people this year alone. Assuming a steady increase in the revenue per employee at the current rate of 10% per year, and assuming a steady increase of 25,000 employees per year, Infosys will be making $8.3 B for 2010-11 employing 170,000 people.

Total output of IT-ITES will be approximately $90B by 2010 (I am being optimistic compared to NASSCOM’s projections), where Infosys will be contributing 10% of it.

While this trend is encouraging, it is not good enough. I keep saying this- we need to look at better and efficient ways to bring in money into this country. Those companies need not be confined to electronics, computer or IT related. One could look at other companies that are based in technology and research. For example, Biogen Idec, a US based healthcare company which makes discoveries in therapies make $2.4 B with 3,300 employees. That’s approximately $700,000 per person.

Wednesday, April 11, 2007

Drivers for penetration of broadband in India

As I discussed in the previous topic, the drivers for penetration of broadband in India will be-

# Decreasing cost per line
# Decreasing operating expense
# Decreasing cost of PC (or similar device)
# Social attitudes and habits embracing broadband facilities
# More Indian content

Decreasing cost per line

The cost per line right now runs between Rs. 7,000 and Rs. 10,000 for DSL services. However, it will be much higher for WiMAX in 2007, and comparable in 2009. Till 2009 the major customers for broadband from WiMAX will be enterprises and SOHOs. The residential broadband users will not contribute much to the incomes during this time. However, the revenues from enterprises and SOHOs are substantial and are a market worth pursuing; hence we will see some deployments. Residential broadband services will start turning out to be lucrative only in 2010 when the existing network can easily be used for increasing the capacities to cater to homes. That’s when there will be huge change in the deployment models by the operators who will vie for the residential customers. Each operator announcing a better deal than the other will push down the prices and also increases the subscriber base.

The decrease in cost per line will come from two factors- decrease in the cost of equipment and from the inherent advantage of wireless when adding new lines. The base stations which cost $3000-$5000 now will start costing $600-$1000 by end of 2009. The CPE (Customer Premise Equipment) which costs $200-$400 now will start costing $60-$80 by end of 2009. The cost of a wireless network is high in the first stage of deployment, because of factors like taking up space for tower, erecting the tower, cabling, housing, and connecting the tower to the network, installing base stations, etc. However, once those costs are recovered from enterprise services, adding new subscriber will come at a minimal cost. In comparison to a wireline network where each additional subscriber may cost more or less the same, in wireless network, each additional subscriber will be minimally higher.

Decreasing operating expense

Look at who is going to provide the bulk of broadband services in India. They are all major operators who already have cellular and landline networks. While laying out the WiMAX networks, those operators will combine the operations with the existing operating premises thus incurring marginal increase in costs while deploying and managing wireless broadband services.

We will see more and more infrastructure sharing between various operators which was completely absent till now. Especially, in rural networks, this infrastructure sharing will turn out to be mere common sense. The operators who own the spectrum will resell the spectrum to other smaller players who will mushroom in various parts of India to cater to non-metros. The towns and villages may not be very attractive to some operators right away and they may like to sell this franchise or sell these frequencies to smaller WISPs and operators. Also, the coming of Virtual Network Operators will add to competition.

Decreasing cost of PC (or similar device)

PC (or a device similar to that) is already becoming cheaper. With projects like one laptop per person, and other initiatives, the cost of PC is going to be less than Rs. 5000 by 2009. Other concepts like Novatium, if they tend to be aggressive can also make a marginal impact on the penetration of broadband. Companies like Intel and Microsoft (or will it be Linux?) will need to come up with exclusive strategy for catering to this revolution (and I am sure they will).

Social attitudes and habits embracing broadband facilities

Every school in India is going to have a PC and also an internet connection. Those kids who never knew PCs will now be used to using them at these schools. These kids in turn will make way for their families owning a PC and therefore a internet connection. All government employees are now moving towards using PCs and broadband connectivity. Those people who have never known PC or a broadband connectivity are now moving towards embracing these technologies. No family wants to be left behind when it come to his/her finishing up his/her homework using Internet at home. Some of these social changes and habits will increase the penetration of broadband.

More Indian content

What will further fuel the penetration is the content which is more relevant to the masses of India. Subscribers would like to see more content tailored to their needs and desires. More Indian language content, more applications suitable to Indian social context will come up in the next few years which will in turn contribute to increasing broadband penetration.


[12 Apr 2007] Manoj Kohli, President and CEO, Bharti Airtel, has this to say:

"All I can say is that by 2010 the market is estimated to have 400-500 million subscribers.

We are looking at a market share of at least 25 per cent, i.e. 125 million subscribers. We have stepped up our Broadband penetration plans and will continue to lead the wireless market share with our passion to deliver the best to our customers."

[12 Apr 2007] In a bid to catch them young, Bharat Sanchar Nigam Ltd (BSNL) is rolling out its broadband services to one lakh schools across the country by December 2008, as part of a project being embarked upon by the Ministry of Human Resource Development.

[03 Jan 2009] BSNL, Novatium launch low cost PCs.

Monday, April 09, 2007

Broadband Revolution in India

[This is based on the presentation I made in Barcamp Bangalore 3, March 31st 2007]

Mobile revolution

We have seen the Mobile Revolution in India. It is still happening. There are enough indicators to suggest that Mobile penetration will exceed 400 million subscribers in India in the next few years. Let’s pause and go back a few years, say 2001. Could any analyst predict that we would have 200 million subscribers by end of 2006? Could anybody in the industry predict that we would be adding more than 5 million subscribers a month?

The growth of mobile penetration is mind-boggling and is quite dramatic. It came about because of some of the factors listed below. Note that it is always easy to look back and analyze why it happened. That’s what I am doing here.

Key Factors for this explosion

* Late start
* Cheaper equipment

* Pent up Demand

* Innovative and Bold Deployment Strategies
* Population

First, India embraced cellular when 2G systems were already deployed in most parts of the world. Having completely skipped 1G because of late start, India went straight to the superior 2G systems. Most of the lessons learnt by the European deployments could be transferred to India. Second, India embraced GSM nearly eight years after it was taken up in Europe. Most of the equipment had already become very cheap by then. This allowed for mass deployment in the country. Third, there was huge pent up demand for basic connectivity which was not served by fixed-line telephony. Fourth, India embraced certain bold strategies- like that of Airtel which has outsourced complete network deployment to Ericsson and management of networks to IBM; it has introduced pre-paid subscription like no other country. However, a note of caution here- they were quite stupid not to have implemented sharing of tower infrastructure right from day one. They seem to have woken up quite late on this. Fourth, with a billion people anything you do seems to pick up volumes. All these factors put together has resulted in mind-boggling and dramatic mobile penetration in India.

State of Broadband in India

Take a look at current broadband penetration in India? It’s a mere 2.5 million subscribers (or less). That’s less than 0.25%. Most indicators of technology penetrations, such as telephone, PC, mobile, broadband, have a direct correlation with increase in the GDP and per capita of a nation. However, India’s broadband penetration seems to be extremely low. Is it that we are going to skip broadband the way we skip 1G and the industrial revolution?

The penetration of broadband will increase in the next few years and will catch like a wildfire when suddenly the cost of adoption and network deployment and maintenance will turn out to be extremely low compared to the kind of market demand it has.

The drivers will be-

# Decreasing cost per line
# Decreasing operating expense
# Decreasing cost of PC (or similar device)
# Social attitudes and habits embracing broadband facilities
# More Indian content

Broadband Revolution in India

In effect, I believe that there is a very big room for growth of broadband penetration in India. With the decreasing cost of PCs to under Rs. 10,000 and then to under Rs. 5,000 soon, and with increasing in content for Indian masses, the broadband penetration will be going through a revolution, and I call it the Broadband Revolution in India. The cost per line will dramatically reduce from the current Rs. 7000-10,000 per line to around Rs. 1,500-2,500 per line by 2010.

The pieces of puzzle are falling into place. With advent of wireless broadband (such as WiMAX and WiFi), with decreasing costs of PC, we will see the penetration grow slow and suddenly, when the price points have achieved that critical milestone, it will take a dramatic upswing and go on an exponential path for the next few years. In my estimation, by end of 2011, Indian will have more than 35 million broadband subscribers and by end of 2013, we will have nearly 1 million subscribers.

My question to all of us

Indian telecom operators are smarter than the rest of us. They usually wake up quite early to realize the potentials of Indian markets. They will definitely get themselves geared up for this oncoming Broadband Revolution. However, will the Indian entrepreneurs, Indian telecom vendors, Indian VCs, Indian startups wake up to this? When Indian Mobile Revolution happened, it was foreign telecom vendors which benefited. They supplied the radio access network equipments, and they supplied the core network equipment. They also supplied the mobile handsets and PDAs. Indian telecom operators had no choice but buy equipment from these foreign players.

Operators like BSNL, Airtel and Reliance throw open tenders worth billions of dollars, and most of these monies are taken up by foreign companies. Almost no domestic company seems to wake up to capture some of this market share. We could give ourselves an excuse that Indian ecosystem was not conducive to create such suppliers in India. That we didn’t’ really anticipate or predict the oncoming Mobile Revolution to benefit from it. Will we give ourselves the same excuse for missing Broadband Revolution, or will we do something about it?

Friday, April 06, 2007

Happy Jobs

What is a happy job? A ad on TV goes this way:

“Mr. Happy Kumar is SO happy with his job that he can’t see millions of opportunities around him. Don’t make the same mistake.”

In this ad, Mr. Happy Kumar is smiling throughout the ad, he is not tantalized by offers like a big car, big pay check or a big position, and therefore this ad concludes he is making a mistake.

Then there’s another ad from Timesjobs on TV which goes this way:

“Salary is making you feel smaller? For better paying jobs, log on to”

Is there anything wrong with these ads? Or is it just me who find them wrong?

What is a happy job? Most of us, who have spent long enough time in the industry, would agree that a happy job is the one which keeps you challenged and which turns out to be rewarding. We all keep looking for a happy job. If Mr. Happy Kumar is happy with his job, there must be a reason why he is happy. If he is not ready to look at another opportunity, and is smiling all the way to his work and is also happy at work, the job must be quite a good one. Why is it a ‘mistake’ that he is not looking elsewhere?

I would have understood if the ad was a little different. If it was Mr. Sad Kumar who is not happy with his job, and is NOT looking at millions of opportunities, then it would have made sense.

Most of the present generation engineers are turning out to be clueless about what they want. Except for some smart ones, most engineers consider a high paying job as the best job. There is already a big problem with the Indian young engineers, they hop jobs like anything. They are ready to take up a new job even for a paltry increase in the salary. Some of them are not just concerned what kind of work as long it is a high-paying job. Some ‘smart ones’ stay at each job for six months and then hop to another. They keep doing this to keep on increasing their salaries. Added to this is job hopping attitude is their appetite for investments into real estate and automobiles which tend to keep their hunger for increase in salaries ever higher.

Such ads are adding to fuel to the fire. Or are they targeting the right customers?

Is salary the only criteria for choosing another job? Should lower salary make you feel small? Only the guys at seem to understand the real pulse of the young engineers.

Tuesday, April 03, 2007

Barcamp Bangalore 3: My observations

I was at Barcamp Bangalore 3 held at IIM-B during March 31st and April 1st (2007). One can clearly see the interest growing from the first event then to the second and now to the third. The turnout has been doubling every event and on Saturday there were 320 participants.

There was an interesting topic later in the evening on Saturday, delivered by Savita Kini. She asked ‘Are we ready to move from IT services to product making companies?’ And of course, this is a topic of my interest. Soon, it turned into a small debate with different kinds of views being aired.

I want to talk about some of my observations from this debate. First, I want to compliment Thomas Friedman, Nandan Nilekani, and all those analysts, writers and industry experts who have created a good myth surrounding our IT industry. They have done an excellent job. Because, there were some quite experienced industry folks in that debate who vouched that when work is outsourced to Indian IT companies or when MNCs set up offshore units in India, the wealth created by those units actually stays in India.

I want to dispel this myth here.

How do offshore MNCs and IT services companies create wealth?

When a IT services company such as TCS gets a job from company ABC, it pulls up N number of engineers, and what TCS gets in revenues is N times a certain fixed amount per year. Therefore calculating the revenues of any IT industry thus becomes very easy. The average $ per person is usually known – and it hovers between $30,000 and $40,000 (for top IT services companies). So, if Infosys has 60,000 employees, then its revenues should be $1.8 to $2.4 Billion. The same rule applies to almost every company. A good services company with 3000 people would be making approximately $90 Million. No additional revenues come from the actual sale of the product. Say, the product is sold 10 times or a 100,000 times, the TCS would still only get its revenues from the number of employees it contributes to servicing that product. None of the profits from actual sale of the products ever reach the Indian shores. The revenues (and the profits) from the sale of actual products go those countries where these companies are housed. (To US in case of Intel and Microsoft, or Finland in case of Nokia).

When a typical MNC puts its offshore unit here in India, the only motive is to cut down the costs. While going for offshoring, they decide on the non-critical work and send it to its offshore unit in India. If an engineer in US costs $120,000 to the company, his typical cost is less than $40,000 in India. Hence the cost advantage! However, no revenue that comes out of the products developed or services rendered by these offshore units actually come back to India. If ever they do, they only come back in increasing the headcount of its offshore units. When Nokia sets up a unit in India, the Indian unit will not generate revenues because of the units sold, but only because of the money pumped in to staff the engineers.

To think or believe that we are creating the wealth the same way those product making companies are creating is a myth that needs to be shattered.

In the colonial times, Indian babus were used by British to act as middle managers to the British bosses. The British, by paying a good salary to Indian babus, above that of the slaves or the natives, were able to keep these Indian babus happy, while they continued to reap the benefits of the toil of the workers. That’s why Indians were taken to West Indies, East Indies, Africa, Fiji, etc. The Indian elite babus who quickly learned English, Mathematics and other desired skills became extremely good middlemen (managers) who served British to help them continue their dominance over other populations.

Second, a lady asked if NOW is the right time to start product-making companies, especially since Savita listed so many challenges on her discussion board.

Shouldn’t we wait for the opportune time before we start product-making companies in India?

Frankly, I believe that any kind of analysis should be left to consultants, advisors, bankers and VCs. Entrepreneurs should just stick to DOING it. There’s no opportune time for doing anything, not even for war. By the time we in India actually solve all the challenges that Savita listed (which included like Education, Discipline, Mindset, etc), it will take an eternity. We don’t sit down solving the barriers and then go about doing it. We just do it and in the process hope the barriers go down. We are already late according to me (and Savita and few others). While Indians celebrate the fact that we are going to be the services hub the way China is now the manufacturing, thereby implicitly shrugging off the responsibility to foster other industries, China is already transforming itself into a global product making player. Haier, Lenovo, ZTE, Huawei, etc, have already arrived and many more are in the making.

According to one observer there is no market for products in India. Some of us had to cite some examples to wake him up to reality. That India adds 5 million subscribers in India per month. We need to look at some more examples. That India is adding thousands of cars per month. That India consumes food, clothes, electronics, etc, with great appetite. Isn’t it surprising to assume we do not have the necessary market when ever top brand of the world is making inroads into India to sell Indians goods and products?

According to some people participating in the debate, India does not have brands like Nokia and Microsoft, and hence cannot enter the global arena. The question is do we really need brands to enter the products-market? Does the brand come first and then the product or the product first and then the brand? While we Indians keep debating if there is a market or not, ZTE (China) is entering India with its own low-cost mobile handset and making inroads into India.

Third, when I said that one of the Indians attitudes is not going all the way, which I call last-meter problem, someone suggested it was the same with Microsoft.

What is the last-meter problem of India?

Look at an Indian road, the last-meter is usually left unpaved, and this seems to cause lot of damage to the road. That last meter produces all the dust, and is also the culprit during the rainy season. The erosion will eat into the paved road itself and very soon we need repairs. This seems to be the perpetual state of affairs in India- start- incomplete- repair. Not finishing up the job to its absolute completion is something like a disease that is all pervading. Half done, incomplete and shoddy jobs are seen everywhere. I call this the last-meter problem. We don’t go all the way; we seem to run out of gas during that last meter.

However, to say that this is the same with Microsoft is far-fetched. Agreed, most engineers seem to detest Microsoft for their monopoly. But, we have to admit that their products are definitely superior (to a layman user) in comparison to anything out there in the market. While an engineer may marvel at the open source and other competing products, Microsoft, by pursuing the common layman as its user has made excellent contribution to penetration of PC and hence Internet, and hence to dissemination of knowledge to every corner of the planet. For an idealist or an expert engineer, Microsoft may seem to be flawed or not perfect, but to an ordinary user its closest to getting the simplest software to deal with an otherwise complicated PC. The last-meter problem of India is not analogous to imperfect Microsoft products.

Overall, this barcamp was great. I got to meet some interesting people and its always fun talking about some of the topics that interest us the most.