I have a different perspective on this, having built a product company (Zoho Corp) in India for the past 13+ years. Basically India's emerging IT industry coincided with the global credit bubble of the past 2 decades, which simply inflated the IT services industry to monster proportions, in the process sucking the oxygen of talent out of the system. So product start-ups that could have been never got off the ground.
This is the mirror image of the process in the US that funneled talent towards the financial complex, which meant the IT industry had to import talent or outsource the work. A good percentage of the talent came from India, and a lot of the outsourced work went to India. In that environment of easy money flowing into services, product companies found it hard to recruit talent (which the author of the post also alludes to).
This process basically derailed the emergence of product-based IT industry in India. Some of the Indian services giants of today (Wipro and HCL in particular) were product companies originally. They built some innovative products in India (an x86 based Unix well ahead of Linux, an IP concentrator/router for ISPs in early, to quote two examples) in the late 80s/early 90s. But once the 90's boom got underway, they found it far more lucrative to rent out the talent than to build products.
In Economics, it is impossible to argue "what might have been", but I believe the configuration of IT industry in India points to how the bubble massively distorted an industry.
"an x86 based Unix well ahead of Linux, an IP concentrator/router for ISPs in early, to quote two examples"
Any links ?
Other than the distortions that 'easy money' introduced, the few startups that tried valiantly to make products never even made it even to the foot notes. Some of these guys were so idealistic that they committed themselves to a non-existent "Indian market". But there is easy money to be made in the US as well.Dharmesh'sixth point about cultural factors might make people uncomfortable, but it's generally true of the IT folks of the last two decades:
"I’ve found Indians to be almost overly practical (in the short-term sense) and not passionate about some of the softer things (like user experience, marketing, branding and other things) which in today’s world are large contributors to future outcomes of software startups. "
These same reasons generally apply to Chicago, Atlanta, Philadelphia, Los Angeles, San Diego, Omaha, Manchester, Paris, Munich, Tokyo, Osaka, Beijing, Hong Kong, Sydney, Melbourne, Accra, Sao Paulo, Buenos Aires, and every other major city that asks "why not us?"
To do a product company, you need 1) talent, 2) money to ride out the non-revenue-generating product development, and 3) precedents to help you navigate legal/vendor/cultural issues.
The two biggest pure software startups (37signals and Matasano) I know in Chicago got around this by "being Indian" - starting off as consulting companies. Even Zoho started off as an offshore consulting company for many, many years.
On Zoho, I want to clarify that we always were a product company, with product development based out of India from the beginning. We never followed the traditional offshore consulting model, which we felt would make it difficult to build products (PG has written about the difficulty transitioning from consulting to products).
We started the company selling software to networking OEMs, 1996-present, I still do sales calls for this software, we are the leader in that niche (http://webnms.com ). From that we expanded to build software that manages corporate networks, 2004-present, we are doing quite well in that market (http://manageengine.com ). The latest evolution is Zoho, 2005 - present, it is our most well known division, so the company changed its name to Zoho Corp this year (http://zohocorp.com ).
In silicon valley these would have been 3 different companies, with the first 2 "exiting" (mostly by being acquired). We preferred to keep building.
Sridhar, I had heard (from a relative in India) that the networking products (your first two on your list) were consulting projects that you productized. It appears what I had heard was wrong, and I apologize.
Kudos to you for building a product company from day one; that's the tough road, and I commend you for pursuing it.
This is plain wrong today. As a person who started a company in Hyderabad, India, I know that the state governments of both IT Hubs of India(Hyderabad, Banglore) try very hard to be non-beurocratic. To register a pvt ltd company, I gave my CA, 25000 INR (~500USD). I never had to go to any govt office, they bought in some papers which I signed, and that was all it needed to get all required paperwork etc. My bank sent a representative to my office to open a bank account, and I didnt even have to go to my bank to open an account. Admiteddly all this took about 2 weeks, but I never felt it was bureaucratic.
Anyway this point is moot if you are talking why no product companies, as this impacts any type of company equally.
India follows the "Sheep Herd" mentality.
The whole country's economy is based on people getting into "Profitable" domains mostly following the success of a pioneer in the field.
The most recent example of this ideology is the "Business Process Outsourcing" industry.
New BPO units are propping up here and there at a dime a dozen leading to a quality deterioration in the final deliverable.
This process will continue till a saturation level is reached and then they will wait till another "Killer" domain picks up momentum.
Till then India will be in a so called "Calm Period" where nothing great and major takes place.
Until recently, the fact that the worst Java-speaking engineer in the room got a 50% yearly pay raise and that annual turnovers among engineering staff were close to 50% rather sharply limited the desire of technically inclined founders to start companies and their ability to retain talent.
In the last three years, the company my day job works through has burned through, oh, call it 5 PLs on the team that I manage, and about 100 engineers. Fully staffed strength for that team is below 30.
This is a constant bugbear for us and annoys my company more than every other issue we've had with India save one. (The biggie, hopefully not applicable to startups.)
India is still far behind in terms of infrastructure( bandwidth, payment, IT penetration ), very few people have made business of into fullfilling indian market, one of the few who made it were tally( indian accounting ) and naukri( job site )
IIT and IIM alumnis mostly prefer to go with business plans and want easy money fro VC's which does not happen very often
Most of the other so called products,services are either have US focus or try to clone existing products for Indian market
We at RailsFactory / Sedin Technologies would like to be in product space, but we are into consulting to build our expertise and network with people and understand end user usage for various services and products we build for our clients
So, you need several local-state-government sponsored ycombinator clones, focused on elimination of 5. 3. and 4.
For example, Sikkim is a great place to start.
This is the mirror image of the process in the US that funneled talent towards the financial complex, which meant the IT industry had to import talent or outsource the work. A good percentage of the talent came from India, and a lot of the outsourced work went to India. In that environment of easy money flowing into services, product companies found it hard to recruit talent (which the author of the post also alludes to).
This process basically derailed the emergence of product-based IT industry in India. Some of the Indian services giants of today (Wipro and HCL in particular) were product companies originally. They built some innovative products in India (an x86 based Unix well ahead of Linux, an IP concentrator/router for ISPs in early, to quote two examples) in the late 80s/early 90s. But once the 90's boom got underway, they found it far more lucrative to rent out the talent than to build products.
In Economics, it is impossible to argue "what might have been", but I believe the configuration of IT industry in India points to how the bubble massively distorted an industry.