Thursday, July 06, 2006

Why Infosys can’t be a product making company?

Before I answer this question (not that I have this beautiful and simple answer which suddenly solves the problem), let me talk about how we got here. Infosys is a great company. I admire Narayana Murthy for what he has done. He is like a WWI War Hero. He fought the war in trenches, faced the gas bombs and artillery shells. He created a services company and made an empire out of it. The fallout of building such an empire has been great for many other companies as well. MNCs started to look at India for outsourcing and nowadays Bangalore is compared with Silicon Valley itself (though that comparison is more hype than truth). Infosys is a great services company. Nowadays you can almost get any kind of software from this company, banking, finance, enterprise, telecom, satellite, wireless, nanotechnology, etc; you name it you get it. While making this superbly well-oiled machine for services (plus consulting and outsourcing), this company has created many war heroes who can fight a WWI with maximum efficiency and minimal cost.

In addition to Infosys, there are other great services companies, like, Satyam, HCL, Wipro, TCS, etc. The combined intake of engineers by these companies is pretty high. Some of these companies take 60-80 engineers from a single college. In the whole process, they create many engineers who are good at delivering services. Services companies have only two factors to play with- Utilization factor and Average Salary. The mantra is- “Keep most of your assets (employees, labs, and other resources) billable and keep the average salary of the employees to minimum”. Give this as a problem to any lay man and he will come up with almost similar strategy most of these services companies employ. You want your engineers to be on a project all the time. So, you scout for projects in all domains and put your engineers to work on them. Though a certain level of skill/expertise is required, you also figure out that most engineers can take up almost any project and work on it. You create engineers who are jack-of-all-trades. Then you also have a set of excellent managers who manage the gargantuan task of delivery (in time and with quality) using these jack-of-all-trades engineers. To keep the Average Salary low, one has to continuously recruit fresh engineers and move the experienced people out or to a management role. So don’t be surprised to see five-year experienced engineers becoming project leaders, and seven-year experienced engineers becoming project managers. Over a period of time, what you get out of such companies are excellent managers; managers who are proficient in delivery mechanism, who know how to thrive and deliver in a services environment, where the parameters for success are very different from that of a product making company. Those who do not succeed in such an environment will be weeded out slowly; or they force themselves to fit into such environments by working on their weaknesses and strengths to adapt to such environment. Give this process twenty years and what you have is hundreds of managers and leaders who are good at running services companies. This is like creating many Generals and Soldiers suitable for WWI. (Just for the sake of argument) Imagine WWI to be services industry while WWII is a product making company.

Now, how can these leaders be suitable for product-making industry (WWII)?

Just by infusing cash and incentives you can’t make product making companies out of this environment. To wage a new kind of war these Generals and Soldiers have to consciously adapt to new rules. Yes, they have the ability to adapt but will they do it? Why should they do it? Especially after winning the laurels and successes in WWI they are keen on keeping the warfare more WWI-like and reap rewards. It is not in their interest to change the rules of the game. Why should one change the rules of the game and turn out to be a loser?

Therefore, all the success stories we get to see are come from services industry. The ex-entrepreneurs are all from services companies and they tend to promote more services company. Books are written on how Bangalore is an ideal place for outsourcing, off-shoring and cost-cutting (by giving projects to services companies). The MNCs who open up shop in Bangalore tend to look at Bangalore as a cost-cutting center, relegating the kind of work they would have given to other services companies. The VC firms who only see such success stories end up promoting services industry. The partners at VC firms have been on board of such services industry and have now attained a flair for advising and helping such companies. Their affiliations and networking is in that industry and hence add value for such industry. Some of the ex-entrepreneurs who have joined this VC firm are comfortable in measuring, assessing and adding value to services industry. If you want to fund a startup that has big names in the industry, those big names come from services industry. The community of VCs and and ex-entrepreneurs consists of people who have waged this services war and have succeeded in this game. They tend to promote more of such industries. Familiarity is the key aspect. It breeds the same kind. So, what we have is a self-perpetuating environment, also called ecosystem, that promotes services industry. There is no room for product-making companies here.

If there are any product-making companies out there, they are exceptions. Such startups may find it difficult to get senior team members with product-making experience. Most of the potential team members with 8-12 years of experience are now managers in a services company. Are they valuable to a product-making startup? Even if they are, will they leave their cushy job and lifestyle to take on this new kind of warfare which they are not used to? Moreover, the startups in product-making space will not find advisors who are veterans of this industry. The VCs are not ready to fund such high risk companies who do not have enough exposure to this different game. Why should they when they have a choice to start another services company? Therefore, most product-making ideas do not take off.

You need precedents to create such an environment but how can you create such environment when you can’t even start off? That’s where it becomes very tough for someone to startup a product-making company in India.

Another analogy that I use is that of a space probe (like Voyager) launched into space. You have two choices. You can move to a planet like Jupiter and use its gravitational force to become a satellite of Jupiter and stay there forever, or use the same gravitational force plus an added thrust to move to the next planet. Most of the Indian companies tend to choose to stay at one level, that is services company, and do not want to venture further to the next level, that is product-making company.

Why do I stress so much on product-making companies?

It’s simple actually. [To repeat myself] In short, it [services-company] is not a scaleable business. If a services-company does $2 Billion revenues with (say) 50,000 engineers, to get to $20 Billion in revenue they need to hire 500,000 engineers. It’s not very practical.

Compare this with a product-making company. A small group of individuals can make a big difference and if successful can create mega business. And it’s not hard to think of a 50,000 people product-making company with $50 Billion revenues.

[Note that I took Infosys just as an example. I could replace it with any other services company and this topic still holds good]

24 comments:

Rajiv said...

There is a big trend towards products in India. A lot of the consulting houses are creating their own products, most of them for in house consumption (eg TCS's own project management tool called IPMS). The ambition is to market them on a global scale one day and more importantly learn the product development cycle.

On the other hand a lot of product developemnt companies have moved their R&D to India as well (eg. the erstwhile Lucent). The environment and work culture is quite differnt from the services companies there. The generals as well as the trench soldiers are getting the exposure to product development.

In the meantime India is losing its cost benifit with rising salaries, increased cost of living and new emerging economies with an english speaking cheap labour. It will take time but is a real and emerging threat to India's IT economy.

We simply have to move to making products. We just cannot stay in our cushy jobs and wait for the inevitable to happen. We learnt how to handle clients. Nobody taught that to us. Similarly we'll have to apply our knowledge and create products.

The tipping piont is apporaching.

http://www.alwayson-network.com/comments.php?id=15024_0_38_0_C

Democratic Geek said...

Good article, but you are missing a very important point. In the 21st century, manufacturing is not going to be the driving force, it's going to be the services sector that drives the world economies like India. USA, for example, is transforming from Manufacturing to Services economy. As a lot of economists etc have talked about, India is leading right now because of the outsourcing and offshoring services it provides to the world. I agree we are not huge, but if the growth rate is 30% annually for Infosys, Wipro etc, they are certainly going to make a difference. R&D Services (Google, Microsoft and other IT companies + other sectors like Pharma) are also playing an important role. If it weren't for services that India can provide, why would SERVICES companies like Accenture hire 10000 people annualy?
Also, only those products are good that have some competetive advantage. As per software companies, there just aren't enough such companies to compete with companies like Cisco etc. If there's a new innovative product, it is just bought by big companies like Microsft and Google these days for web products, which then take care of it.
Companies like Infosys have had good strategies too, and the reason they are operating from countries like China and Latin American countries too to take care of rising costs in India.
If you want to see Infosys to get to $20Bn mark, it is not with the IT sector, we need better economic policies, which thanks to our communist government, is not going to happen.

Mahesh.R said...

Intersting post. Though it is true that most of the companies making money now are services company. To a large extent these services companies have blossomed because MNCs have given them such jobs and these companies have enough talent to take up the job and do it proficiently. First India used to get only services jobs. The trend is on the change now. More and more companies are setting up research facilities in Bangalore, Micrososft and Google , for example. IBM have been here for sometime. The MNCs are spotting the immense amount of talent in India and are recognizing that they can be used in product development. With them building the base, it is upto us to slowly get rid of the "services only" tag to "services and products". It is a matter of time that more VC's will come foward to support good ideas of product development. As you yourself said, there is more money in Products than in Services.

CrazyEngineer said...

Intersting post. I have a simple funda: If a product takes 10 bucks to make, you'll sell it for more than 10 bucks. But for a service that is worth 20 bucks, you will be paid 10 bucks. How about that?

Look at the whole thing from India's growth perspective. Our services industry is totally dependent on those product making compaines in US. Narayana Murthy says "We do what we are good at".

The big question is, where will we end up servicing the others.

Have a look at -

http://crazyengineers.com/me/?p=30

-Crazy Engineer-

Sujai said...

Dear Democratic Geek:
I don’t profess that I am an expert of the industry, but I know the industry enough. I don’t think I am missing a very important point. On the other hand, I conclude that you did NOT understand what I mean by product-making companies. You seem to be of the opinion that “product-making companies = manufacturing industry” which is NOT correct.

Let me illustrate (lest others think on similar lines). Companies like, Nokia, Ericsson, Cisco, Oracle, Microsoft, etc, are my examples of product-making companies. They are not necessarily manufacturing all their products. They may be outsourcing manufacturing to other companies. Many Chinese and Taiwanese companies take up these manufacturing services assignments from these product-making companies. On a broad-level (just for this argument), one could divide this technology world into product-making companies, manufacturing services companies and software services companies. Most Indian companies come in the third category- software services. I was not referring to manufacturing industry in this topic. [I will write more about this industry later].

>> but if the growth rate is 30% annually for Infosys, Wipro etc, they are certainly going to make a difference.

I do not believe that this growth is sustainable. Revenues of most of the Indian services companies are directly proportional to the number of employees. At some point, this growth will have to taper off. Managing so many people on multiple projects will have associated problems and costs and therefore will stifle the growth- that’s my belief.

>> If you want to see Infosys to get to $20Bn mark, it is not with the IT sector, we need better economic policies, which thanks to our communist government, is not going to happen.

Wrong argument. Why Infosys can’t make $20B is nothing to do with IT sector or with Indian economic policies. Indian government provides many sops and tax-breaks, gives subsidized lands, and other incentives to Indian software services companies (like Infosys). Why Infosys may not be able to make $20B is more to do with the kind of industry it is in – software services. If Infosys can manage a 500,000 people company, hopefully, it can be become a $20B company. Contrast this with a product-making company like Nokia- it employs 40,000 people and makes $40B.

Sujai said...

Dear Crazy Engineer:

You wrote:
>> I have a simple funda: If a product takes 10 bucks to make, you'll sell it for more than 10 bucks. But for a service that is worth 20 bucks, you will be paid 10 bucks. How about that?

How about another simple funda:
Services Company:
Let's say you are making revenues of $1M with 25 people servicing 1 customer. To add another customer to make $2M (or double the revenue with same customer), one has to hire 25 more.

Product-making company:
Let's say you are making revenues of $1M with 25 people selling to 1 customer. To add another customer to make $2M (or double the revenue with the same customer), one has to hire 3 more.

Take this to billions and the result is more obvious.

Anonymous said...

Well one question remains... as you mention services industry require average qualified hard working people... while product companies requires really creative and technically strong people... where are you going to get that in India? Forget about managers, its simply difficult to recruit and retain highly qualified people in India and secondly what about the failure rates of a prodcut company? Compare it with services industry ... A product company can only do well only when the product it developed works wonder in the market... there is little chance of revival if it fails... whereas a services company just need to garner projects.. even if doesnt get one big project... many small projects are enough to sustain it..... its a simple equation: prodcut company = more risk more gain; service company = less risk less gain; and indians are known to be less risky :-)

Ketan

Sujai said...

Dear Anonymous:
I DO NOT say that "services industry require average qualified hard working people... while product companies requires really creative and technically strong people..."

I SAY that services and products are two different games. To have a product-making ecosystem, you need lot of people who are experienced in one domain for a long time. I believe switching domains frequently may not create such experts.

I WILL NEVER SAY that one industry requires less "creative" while another industry requires more "creative". If my article suggests such an impression, you have to forgive me.

>> where are you going to get that in India?

There are many companies in India where there are people with good experience who are seeking out opportunities so that they can create products. These companies include both services companies and MNCs. The challenge is - will they risk?

Democratic Geek said...

Dear Suraj Karampuri,
It is not true that employment is proportional to the revenues we can get. You might have 50000 employees, get the same contract from the same company, but still get paid less, as Infosys does. The reason for that is Indian companies still don't have that great brand recognition compared to IBM or other services companies.

Look at banks like SBI, ICICI - they are also services sector and not "product-making", but they are making a lot of money. If you talk about same industry in the US, banks like Citibank and BOA are making billions of dollars without "making the products". Although I do agree that product-making companies make more money, but it just isn't possible for India. Moreover, can we even compete at international level yet? Most of Indian IT companies cannot, only a few might be able to like Infy, Wipro etc.

And coming back to economic policies, these are affecting these companies. If policies get better, Infy can do better. Just giving subsidized land doesn't mean that goverment is helping a lot. If the government imposes employemt quota based on castes, then it's going to hurt India Inc, and which MNC would want to start their operations in India knowing that they will have to employ people from OBC and SC/ST.

To double the revenues, you don't have to double the number of employees. Instead, you have to be able to offer better services. In services sector, it's more about the value-added stuff and goodwill of a company than the actual services. That's the reason that companies like IBM and Accenture can charge 30% more than Infosys and still get the work done at same cost in India. Companies pay higher premium just because of the reputation of these companies. Infosys is already on it's way, but it will take time.

The new concept of global economies is not about concentration on product making or services, but about concentrating on what someone's better at. India is better at services sector, and not product-making like cell phones etc. India is becoming a services-providing company, with 50% of GDP coming from Services sector. You also need to remember that companies like Infosys, TCS are not just Software Services now, they are transforming themselves to become Consultancy companies, in order to compete with global giants like Accenture which are more of consulting companies than just IT companies. This will definitely help these companies in India.

Sujai said...

Dear Democratic Geek:
My examples for software services companies are NOT SBI, Citibank and BOA. Their clients are people. My examples are Infosys, Satyam, TCS, etc- their clients are in turn the companies like SBI, Citibank and BOA.

India cannot be complacent that it is only good at software services and hence conclude that it needs to concentrate only on software services. This business is untenable. Where is the rest of technology ecosystem which can sustain these industries? They may want to climb up the ladder and get that 30% more (like Accenture), but then they need to show that kind of experience and expertise which they lack (and which they cannot build given the nature of the parameters they work on).

Even countries like China are getting into the mode of making their own brand products- Haier, Huawei, Lenovo, etc, leveraging their manufacturing experience.

Amit Arvind Badiyani said...

At a broad level, I think there are very valid socio-economic reasons for the things to be the way they are. Firstly, we as a country are still poor people. We are still striving to meet our basic needs like roti, Kapda, Makan, bijli, pani, sadak, telephone etc. and easy money offered by services helps us meet these needs soon.

Another factor is relatively less acceptance of failure in our society which discourages us from taking bigger risks of trying to develop and sell products particularly, when we have the relatively safe option of making money by selling services.

Third point is the latest trend of offering even products as a subscription based service. This works to be a win-win situation for the product vendor as well as the customer where customer need not worry about upgrading product from older version to the new one and the vendor need not worry about IPR related issues. This is a less risky option for both the ends in the transaction.

dazedandconfused said...

Agree and disagree. Yes, product making companies are more profitable, if they are succesful. For every Apple, there's a Netscape. Bigger risk, bigger returns.

Making products in the IT space is at a higher level of maturity. Many Indian companies are not there yet, though there was i-flex, and Infosys has a product called Finacle, which is doing quite well.

I think Indian companies should continue to concentrate on Services. That's where we are good at. I don't agree that it's not scaleable. Why can't Infosys get 500000 engineers working for them? That many graduate right now every year, there might be more in 10 years. And then who says that Infy needs to only recruit from India?

ankurg said...

doubting the capabilities of the existing services IT companies is something i will not comment much on, but the comaprison of revenued to employees is something i agree with.

lets look at on e product company called "SubexAzure" ( which i am an employee of ).

Infosys with an employee strength of 50,000 generates revenues of $2billion so that is $40,000 per employee.

Now looking at the last year results for product subex it generated a revenue of $40 Million with an employee strength of 300 which is $130,000/employee.

with acquisition of another company Azyre solutions this year , the revenues are expected to be $80 Million with an employee strenth of 500 by the end of financila year 2006. Which is $160,000/employee.

This advantage in product comapnies over services companies comes from the way the two types are operationalised.

But as also mentioned before this is correct to say that the risks are higher and the software product that is built needs to be very successful to achieve a significant success. Which did happen in case of subex with its initial launch of the products and whihc might not have happened in some of the companies and hence we dont know about them.

So high risk high gain is absolutely correct about this industry which keeps most of the industry people and VC's at bay when they have the option to choose services.

(Figures quoted above from www.subexazure.com)

ankur

Sujai said...

Thank you, Ankur, for your statistics. Subex is a great example for product-making companies from India.

Anonymous said...

Hi Sujai,

I agree with your assessment. The DNA of services companies are very different from DNA of product companies. It is tough to make a gazelle out of a gorilla.

Cheers

Sanjay

Dweep said...

Hi Sujai,
Thanks for commenting over at my blog on Wipro and prospects for Indian business. I've responded to your concerns at my blog.

I would like to put in perspective your emphasis on product-making in India.

First, product companies also have scale limitations. Second, you ignore that the largest companies are, in fact, service companies (think IBM, McKinsey, Lucent). Even telecom manufacturers have been moving from pure product development to a mix of product and service offerings (e.g. Ericsson), and IBM is the clearest example of this.

Product development is no silver bullet. In fact, both service and product development require innovation - just of a different kind. My response has more details.

Sujai said...

Hi Dweep:
I responded your comments at your blog.
Sujai

Some of those snippets are here:
Dear Dweep:
According to me Lucent, Cisco, etc, are not software services company the way our TCS, Satyam, Infosys are modelled. I do emphasize that we need to have companies like Cisco - which drive technology and products.

Services companies like Citibank, Vodafone, etc, serve individual consumers and enterprises. Companies like TCS, Infosys in turn serve customers like Citibank and Vodafone. There is a big difference.

McKinsey, Deloitte, etc, can prevail and sustain in only those geographies and economies where there is need for their services- which happen to be high tech manufacturing, large scale production and technology driving economies.

IBM has the required competency and experience because of its rich background in creation and driving of technology segments. It does not come in the same league of other Indian software services companies.

My emphasis has been on technology driving and product making companies. And I emphasize on creating ecosystems. McKinsey, Charles Schwab, etc, feed on such ecosystems. I believe Indian software companies like Infosys, TCS, etc, do not create such ecosystems. While, Tata, Mahindra, etc, do create such ecosystems. With good policies, Indian Defence and ISRO can create such ecosystems.

In my blog, I gave reasons why a company like Infosys cannot transform itself into creating such an ecosystem.

I don't think making products is sexy (unless of course they are creating more Aishwarya Rai s ;). Its is nothing to do with glamour either. It is the ability to drive technology and reap its rewards. Who is reaping the rewards for an innovative product like Apple Ipod? Apple or taiwanese manufacturers? [and who is benefiting from trickling revenues?]

Please look at some of the examples I gave in comparing how our Indian software services companies make revenues compared to companies like Microsoft, Google and Nokia (they are all technology drivers).

Balu Natarajan said...

I work for a product making company in Silicon Valley. I would like to add to what sujai already said. We need to see an attitutde change among service companies. Products are much more risky than services. In services you serve a customer who has already signed up & so you Bill him to make ur buck. In Products we look for customers after having a product ready. if you are not cutting edge ,you would probably loose the race in products while in services non-sense can be fixed and is tolerated .

There are problems with product companies in Silicon valley as well. Poor Engineering efficiency (in terms of execution ), High costs , Not enough money being spent on scalablity and reliablity of the product. In US what a 50 million dollar VC funding could do , the same could be executed by a small group in India for probably 1/5th that cost. But Its really tough to market and sell a product from India. A good product start up would have its Marketting, Sales team in western world. It would have the core architecture (Engineering) team abroad. It would have Team Leads, QAs and people doing maintanence in India .

Anonymous said...

Product companies make money while sleeping but this is not true with service companies.

Rakesh said...

interesting piece...

I agree with Rajiv, i think getting into service industry is the first step in a long march.

India needs more entrepreneurs and a stronger VC culture to think beyond utilizing the cost benefits of the labor.

Rams said...

Brilliant. Required reading for technically minded nerds who feel suffocated in indian companies - this should explain things @work quite a bit and help take a decision. I posted this to shunya.in. I will try to expand on some of the points that you make some day.

I work for a small product startup in Bangalore run by an NRI guy with
impeccable credentials - ran a couple of startups, went to the same schools that the google founders, and Paul Graham did. But I catch myself muttering every day "You can take a man out of India, but you can't take India out of the man". The same obsession with cheap labour, cheap hardware, cheap everything, micro-managing down to the metal, etc etc.

I am pretty sure after working with these NRI, TiE schmoozing types, that the real Indian product companies(like the truly successful Indian service and manufacturing companies) will be lead by Indians based in India and grounded very much in local culture. It is shocking how little originality these NRI types have - chasing tail lights, cashing in on the latest technical fad is all they seem to be capable of.

Anonymous said...

I agree with the main theme of the post.

Somehow the passion for one's subject - which often drives technological breakthroughs and a lot of killer apps and products is completely lacking in the Indian software industry. For example our contribution towards open source products has been negligible. Now - when some of the open source products are getting into mainstream, startups are setting up shops in India to have our engineers do the testing !!! (Nothing against testing per se - what we are talking about is mundane, routine, mind-numbing grunt work, which India is seen aqs the ideal place to dump on.)

The services companies - as the author has correctly pointed out - have no incentive to change the status quo - so the TCSs and Infys and Wipros will contuinue to remain the sweatshops they are. (not that the MNCs are any better. I work for one and believe me I know !!)

Finally - Rams - I have had some experience of these "na-gharka-na-ghatka" NRIs and I do sympathise with you. They are far more difficult to deal with than "goras" and the sooner we get rid of them the better it is for the Indian software industry.

Chus said...

This is what I think: Infosys and Satyam Hiring and Firing

vijay said...

HCL which was a prime player in the computer industry in the eighties lost out to TCS, Wipro, infosys etc. because they missed this outsourcing ,BPO thing. The explanation given by Shiv nadar for this was that he wanted his company to concentrate on technology, to be the next microsoft rather than go into services. Somehow it did not take off and the company finds itself in a bad position now.

The reason that people are averse to taking risks is that there is no structure in our country which cushion the fall. If it could be difficult for HCL it is almost fatal for small players. Unless we evolve some structure to ensure that the failures are not fatal, this trend of going the easier way will continue.