Rediff.com: Betting To Win : Ajit Balakrishnan

Interview Questions

Full Interview with Ajit Balakrishnan

Ajit Balakrishnan:

Ajit Balakrishnan: Though people find it hard to believe, my goal in 1995 wasn't really to make money. Even today it isn't. It is more intellectual curiosity. I wanted to know where this new technology (the Internet) was going. In 1995, the answer may well have been nowhere. Back then, it was not clear that the Internet was a media. But my instinct told me that it would be. The inspiration to start Rediff came from a business course I took at Harvard Business School in 1989-90. We were asked to study Compuserve, an information services company before the era of the Internet. It used a proprietary networking protocol and one had to use a command-line interface to log on to a time-sharing computer. It had news, chat and practically everything the Internet is known for.

The missing link was Windows, which had not yet been created. So, the user interface was a black screen with blue text. The commands were quite complex and required a lot of keystrokes, dots and slashes. One needed to be half-a-programmer to use it. With the introduction of TCP/IP, and the disaggregating in the computer industry, I guessed that it was time to start a new business, which was based loosely on Compuserve, but with newer protocols and technologies.

 

Ajit Balakrishnan:

Around that time, there was a great amount of debate over which network protocol would take center stage. The two contenders were X400 and TCP/IP. As late as 1989, there was absolutely no assurance of which was going to win. TCP/IP did, and Compuserve eventually died because they did not move to TCP/IP. With the new protocol, networks were much cheaper to build; with Compuservetype protocol it was almost like the old landline switches. The connection needed to be kept alive. If you had 10,000 users online then you needed to keep 10,000 connections alive. While at Harvard, we also saw the first multimedia-supporting browser put on Compuserve. It was platform-independent. It was unlike Windows, which was tightly tied to Compuserve's backend. Every piece of technology became independent.

 

Ajit Balakrishnan:

Absolutely, I agree. We launched our shopping site in 1998. Those were the early days of e-commerce and it never took off as expected. There are a couple of reasons. One, credit card penetration is low in India. Today, there are only nine million active credit card users in India. And, debit cards have not yet been enabled to be used on the Internet. In the absence of this, all of us largely depended on a cash-on-delivery model, which is an inefficient system. Any efficiency and cost savings users get with Internet shopping is more or less annulled by this model.

Two, outside the US - and in limited way Germany and France - auctions never took off.

Auctions normally work for products that have reached the end of their engineering lifecycle or those which are in short supply. The CEO of eBay, Meg Whitman, theorized that since consumers in places like India and China are being introduced to consumer markets for the first time, they prefer to pay a premium and buy something new - not second-hand stuff. I think she maybe right. Retailing is still relatively new in this country. In the US, online shopping is dominated by traditional retailers who sell about 5 percent of their products online. Online retailing is in its infancy in India. Retail folks have to stick on for the next two or three years before this becomes anything big.

 

Ajit Balakrishnan:

The arrival of low-cost airlines has created a boom in online purchases. Roughly, according to IMAI (Internet & Mobile Association of India) numbers, two-thirds of online purchases are tickets. This shows that if you can offer something worthwhile and valuable - something users can't get conveniently and cheaply outside - people will pull out their credit cards and pay.

 

Ajit Balakrishnan:

Those were certainly trying times. We had our IPO in June 2000 and in July everything collapsed. It was like sitting on Noah's Ark. But businesses go through cycles and I told my core team to stay cool. I told them to treat it like a downpour and to stay indoors. The downturn lasted for about three years. As experienced entrepreneurs, we kept everything low cost, something we still do today. We had between $50 and $60 million in the bank when everybody else had zero. We did our share of foolish things - but not as foolish as what others did.

 

Ajit Balakrishnan:

We invested quite early in a talented, mainstream technology team. From 1996 onwards, most portals outsourced technology. I maintain that we can't outsource these areas because they are core to our business. In the last two years, we have had a program to partner with young, strong companies of 10 to 12 people. They are innovative, and come to us with good ideas. We, in return, provided them with an innovation platform. In the US, companies like Yahoo normally acquire these companies. We don't. We rather partner with them and make investments if the need arises. These young companies need more revenue than investment. Once they have revenue they can get VCs to fund them. We have a dozen such companies surrounding us anonymously, providing us with a huge source of innovation.

 

Ajit Balakrishnan:

I know there is a general belief in the tech community that there are early-mover advantages attached to technology. It's generally not true. It could be true if a company can sustain that lead throughout its life. Look at the graphical user interface of Windows that made Microsoft so successful - it came out of Xerox. Xerox couldn't make it work for them. Neither could Macintosh. Bill Gates however had a business design to license the software for cheap. The business design was original - not the technology. Look at Google's PageRank. The methodology has existed in the academic community for at least 50 years. To see whose paper has more value, academics check how many others have referred to it. It's the same concept with PageRank: the more people linking to your site, the more popular it is. Or take sponsored ads on Google.

GoTo.com (now Overture.com) has been using it for at least seven years. It's not the person who does something first that succeeds. A window of opportunity is open for four or five years. Anybody who enters during that period succeeds. Later, the cost of catching up becomes too much. Or the product goes into a stage where innovation is not rewarded. In our case, iShare (a desktop tool that shares files) is relatively early and is proliferating like mad. We got our first million users in four weeks. The next guy to try a similar videosharing site, will have it much harder. First, the novelty's gone and second, we execute it well. User already have Youtube and iShare, there's no need for another one.

Technology is essential in the beginning. For example, when NBC came along, it was the only channel available in color. For the next five years they were the king because they were the only channel where one could see programs in color. Soon, everybody else had color. In the media, technology advantages don't last very long. They proliferate immediately. Point is: the early mover advantage is not sacrosanct.

 

Ajit Balakrishnan:

For the last six months, we have been trying to build data models for every part of our business. For example, for our online shopping, when a product is defined, it comes with parameters: description, price, ID, etcetera. We realized that these data structures were different from those being used for our 'product search'. But when we talk about a product, we want it to mean the same thing across five different businesses. A large amount of effort is going behind this. It is fairly new in India and it is quite difficult to find people who do data modeling. So, we took out our textbooks and everybody is busy reading about it - including myself.

Overtime, this will ensure that maintenance of all these complex platforms becomes easier. Second, we have six or seven platforms. These include a shopping platform with a payment gateway; a content platform for the editorial team; a search function, which is a data-mining platform. We are trying to abstract some of these as common Web services across all platforms. For example, we have defined payment as a standalone Web service. A lot of effort is being put in to architect the platforms on a Web services paradigm. It is challenging because standards are still emerging. But, it is absolutely essential to succeed in the future.

 

Ajit Balakrishnan:

A CIO should be the chief repository of the organization's IT architecture. We are constantly adding new services and features and they need to fit. It is almost like the job of a high priest: to ensure an architecture which works and that everything links within it. The second most important role of a CIO is to be an inspirational leader and attract the right number of people at the right prices and motivate them to keep them here. Achieving and maintaining high levels of quality people is a challenge, we face a shortage of talent.

 

Ajit Balakrishnan:

As far as I know, the reason behind our low ranking is that on a per capita basis, access to either PCs or the Internet is low. Anything divided by a billion becomes small. Another reason is that the Rupee-Dollar conversation rate deflated from Rs 8 to a dollar in the 1980s to Rs 50 to a dollar in the 1990s. This made PCs too expensive to buy because their components were largely imported. Another big problem is language. There is almost no content or services on the Internet in Indian languages. The government has found a committee to address this and made me its chairman. At our first meeting in October, the big conclusion was that it is incredibly difficult to author, store and display content in Indian languages. Primarily because when the fonts were created, they were developed to be stored in an eight-bit format. Unicode, as a standard, only emerged later.

 

Ajit Balakrishnan:

If you look at the 50 Indian newspapers sites, each site uses a different set of fonts.

They'll make you go and download the font. We are trying to figure out how to get everyone to adopt Unicode. The government has funded many projects to come up with apps, with authoring systems, display systems and browsers, in Indian languages. Many of these projects have failed. There is a view that cell phones may be the solution. Cell phones have a technological challenge: their keyboards are still in English. You need innovation in voice recognition. One of the projects we have enlisted is a voice recognition system with accuracy levels of 95 percent. Then, a user can speak into his phone asking for, say, a train's timing between Mumbai and Pune. This will translate into text, search a database and revert instantaneously to the user. If we can do that then we can bring the Internet, data mining and other wonders of technology to one billion people. We are trying to figure out how to manage and fund such projects. We want to complete such projects in the next three years.