NET.COLUMNS: The Business of being an ISP in India

MUMBAI (October 30): Many Indian companies are waiting for the government to open up the Internet access business. VSNL has had a monopoly over the past year to build up a business of about 12,000 customers (dial-up) and a few leased-line clients. Most corporates eyeing the Internet business expect the DoT to open up the segment well before the end of the year. What will happen then? Should you be getting into the ISP business? Are there any other business opportunities? Read on.

An Internet Service Provider (here Service is taken to mean “access”) offers connectivity to the Internet. The ISP will lease a pipe to the Net, set up an infrastructure to offer dial-up shell or TCP/IP accounts, host Web sites and offer connectivity for creating Intranets (if the ISP has a presence in multiple cities). VSNL is the only ISP as of today in India. ERNET offers services for a limited audience: the educational and research institutes. NIC has so far been limited to offering its services to government departments.

When India opens up the market, we will have three options: either to make it completely deregulated, or have the ISP licence available for a fee, or to have bidding for offering a 2nd or 3rd ISP licence. The US is an example of the first case; anyone can become an ISP. The email business in India is an example of the second case. The ISP business in Singapore and China follows the third approach. Long-term for India, the first option is the one which makes maximum sense.


Consider for a moment the investments necessary to become an ISP. If you are looking at just a single city, computing equipment (Pentiums or Unix workstations) will cost about Rs 10-15 lakhs. Each phone line and modem (28.8 Kbps) combo will cost Rs 25,000. Assuming a 1:15 ratio of lines vs dial-up subscribers, to support a base of about 5,000 subscribers, we will need 300 phone lines, implying a cost of Rs 75 lakhs. Of course, this investment can be made gradually. A terminal server will also be needed to support tens of phone lines. A router and leased line to VSNL will be needed. Minimum investment for a 64 Kbps line will be Rs 15-18 lakhs per annum (most of it recurring). To support 5,000 customers, it is very likely that a 2 Mbps pipe will be needed, requiring an investment of over Rs 1 crore. Add to it office space, manpower, marketing, etc. My estimate of the minimum investment needed to start an ISP business in a single metro is about Rs 3+ crores.

If an ISP gets 5,000 customers and assuming a revenue of about Rs 1,000 a month per customer, we are looking at Rs 50 lakhs per month, which is about the minimum to run the business. If you look at the national market, it should be possible to have about 100,000 users on the Net in the next 12 months. Assuming that VSNL will have about 50% of the market, it means the others will need to split the remaining 50,000 subscribers between them offering room for a maximum of 10 ISPs. In reality, I do not expect more than 5-6, with VSNL probably retaining a higher market-share (more closer to 65-70%). The attraction for ISPs is going to be on two counts: value-added services to the customers, and also the long-term numbers in the ISP business. We are looking at a potential subscriber base of 1 million in 4 years.

Let us go into some of the issues into greater detail.


Who are the companies most likely to get into the ISP business. They fall in 5 categories: Email providers, Telcos, Cable Operators, Publishers, and Kirana operators. Email providers have little choice: they have to offer Internet presence, else their core business is doomed long-term. They are the first ones who will get in and will also be the best positioned — as long as they still have deep pockets left. Of the current players, Sprint RPG and Wipro-BT look like being the players to watch out for.

The Telephone companies will offer localised services in the cities/states in which they are licenced to operate. It is a natural foray for them into the value added services segment. They will probably be better positioned to create localised infratsructures which they can “retail” out to smaller players. Since data and the Internet business will be miniscule compared to the voice business, the telcos are not expected to give a lot of focus to this part of the business.

Cable operators are in an interesting situation. Recent months have seen a lot of talk about PCs and computers merging, with WebTVs being launched last week. Cable operators are the best off in terms of the bandwidth going into people’s homes. But their equipment will be needed to be upgraded to support two-way communications. Trials are under way in the US, and a company called @Home is going to be launching a service shortly there. Again, it will be an incremental service for cable operators to offer. The cable market is not yet well organised in India, so it will be some time before they begin to offer Internet access.

Publishers have two things going for them: content and reach. They have information useful for their target audience. But they are more likely to host their information on the web servers than get into the business themselves. A precursor to offering Internet access in most countries is a proprietary dial-in service. Only two publishing houses have offered this in India: Business India and the Dalal Street Journal group. Both have met with only limited success.

The Kirana operators are an interesting breed. In the event that the DoT does delicence the ISP business completely, then these are the ones who will really spur the business. Small, regional operators will offer access in specific localities. Their biggest strength will be intimacy with the customer and a level of support which few will be able to match. They will have hundreds of subscribers (building complexes, for example) rather than thousands. They will lease capacity from the telcos or other bigger ISPs. Over time, as the industry consolidates, they will probably be bought over by bigger players. The evolution will be rather similar to the cable industry in India.

One group quite noticeable by their absence in the above list are the computer companies. I do not expect them to become ISPs, rather I expect them to market the services of ISPs.


The all-important issue is going to be the pricing of the dial-up accounts. In the US, prices have flattened at about USD 20 (Rs 700) a month for unlimited access or about a dollar (Rs 36) for others. In India, we are at Rs 30 an hour for TCP/IP access. Assuming usage of about 2 hours a day or 10 hours a week, we are looking at a monthly charge of Rs 1,200. Given the high cost of telecom connectivity in India (especially international), prices will not be going much lower that what we have them.

As I have mentioned before in these columns, the barrier to growth is the high initial charge (Rs 15,000 for 500 hours). I expect private ISPs to introduce more attractive pricing options, including bundling to get in customers. So, even though the effective rate will not be lower, the entry barrier will be reduced which is expected to get more people in. I also expect flat monthly pricing to come in (with unlimited usage).


How will ISPs differentiate them? Quality of Service, bandwidth to the Internet, uptime are the obvious differentiators. In addition, given that the level of computer penetration in India is rather low, customer education, training and support will play a very important role. The media is and will continue to do the job of selling the Internet. The Internet can provide a reason to own and use a computer. But a fair amount of hand-holding will be necessary for the end-user.

ISPs will need to make their money not just on providing access, but also on additional businesses like hosting and creating home pages and web sites (which will become a cottage industry) and marketing the necessary equipment and software the customer needs to get set up (modems, PCs, Web browsers and authoring packages, etc.) Over time, as content providers start charging subscriptions to their sites, the ISPs can serve as the collection points since they already have the billing capabilities.

Who Wins

The ISP business is a long-term business. There are no short-term returns. However, one winner is guaranteed. VSNL will be in the wholesale and retail business, and as such is likely to benefit most. It has also had the advantage of early learning. To make it a level playing field, VSNL should offer only the wholesale pipes and create the nationwide infrastructure rather than compete with the ISPs at the retail level also. I am not too sure this will happen. The one grey area for VSNL is Internet telephony.

Either way, the Internet business is going to grow. And with it will piggyback the Intranet business, which will be the real opportunity. That is one area which VSNL by virtue of its national presence can give VSAT operators a run for their money. It is something which has not yet been exploited: building private virtual networks over the Indian Internet. Only after this begins to happen will Indian corporates really experience the benefits of networking.

This is not to discourage companies from getting into the ISP business. The Internet business changes very rapidly, and if the government policies can keep pace and offer a fair, deregulated market, then we have a huge business in the offing. Worldwide, companies are structuring and re-engineering key processes around the Net. We in India will be doing the same. The missing link is going to be the ISP. Electronic commerce promises to bring about a revolution in how business is done and money is transacted. To benefit from the future, you need to be a part of it now.

Published by

Rajesh Jain

An Entrepreneur based in Mumbai, India.