Thursday, September 6, 2007

India internet - still waiting



I attended a conference on the Indian Internet sector, organized by IAMAI. Technically, the conference also covers wireless, but most of the discussions focused on Internet. Here are my observations, after discounting for the natural self-promotion that occurs at such conferences:
India internet: not yet.

Growth is decent (subs, e-commerce, ad spend), but no hockey-stick effect yet. Given the small base, evolutionary growth isn’t good enough to quickly reach significant scale. We are not seeing the equivalent of cellphones circa-2003, where the industry exploded (triggered by CDMA entry, free incoming and a $10 entry cost to go mobile). A base of 4-5 million Indian homes have PCs, with new PC sales to households hovering slightly north of 1 million/year. To put this in context, the PC base is roughly comparable to the base of passenger cars (remember, the average car costs 10x that of a PC) and the number of households with an annual income of over Rs. 0.5 million. PCs are yet to penetrate the Rs. 0.2 to 0.5 million annual household income segment, that forms the mainstay of the Indian middle class (this segment has roughly 14 million households).

Take all numbers with bags of salt (including the ones I quote!). There are assorted estimates for subscribers, e-commerce, broadband etc (I heard a range from 10-70 mn subs, with the median at an official-sounding 38.5 mn). Rather than get into estimates, let me say that we need to apply a quality-filter (e.g. not just talk about internet subs, but address hours used, frequency, non-email usage, quality of connection etc). I didn’t see any estimates for subscribers for whom Internet is a habit, with reasonable usage beyond email. I put this at 5 mn, give or take.
Online travel leads the way, especially Railways & Air Deccan. Indian Railways has been selling Rs. 30 crore of tickets monthly on the Internet (over 200,000 transactions monthly). On a relative scale, this is still small – 2% of total revenue, near-10% of upper-class (a/c) tickets (if you assume online bookings are predominantly for these classes). Air Deccan uses its internet-based reservation engine for all bookings and has a run rate of Rs. 1000 cr/year. While all transactions use the internet, most of them are intermediated (travel agent, Webworld, call centre, booking office). They didn’t disclose share of direct internet bookings.
Cyber café’s are the primary mode of access. 60-70% of internet users access the net at cyber cafés. I see this as analogous to the PCO-era in our telecom sector. Think about the sea change in phone usage now that 70+ million people carry cellphones and don’t have to rely on a local PCO to talk to other people. The same will happen to the net only when it moves from the café to the home.

Disagreement on definition of broadband. I was disheartened to see India’s internet pioneers say ’56 kbps is good enough for Indian customers’. Seemed kinda regressive to me! I know services need to be priced low in India to reach scale, but that doesn’t mean customers will put up with crappy service. While we pay a few cents/minute for mobile calls, the service is still world-class.
Bottlenecks – affordable good-quality broadband, payment mechanisms and (perceived) security risk (for commerce). Surprisingly, high PC prices weren’t mentioned as much (I’ll talk about this later). At $25-30 million ad revenues and $150-200 million e-commerce revenues (gross), the India internet market is clearly a small market today. There is also no doubt on the direction and that this would eventually become significant. The main question is ‘how soon’? I don’t have the answer, but here are my views on some of the relevant factors and on what needs to change for us to get there.

Let me start with my definition for ‘significant scale’. My medium-term target would be 30 million households having high quality (over 256 kbps), affordable ( under Rs. 500/month for PC-EMI + net access without download limitations), reliable internet access at home. I deliberately pick 30 million, as that is the number of households who’ll earn over Rs. 0.2 million/year in the 2010 timeframe. There is also a qualitative aspect that I’d like to emphasize. I read a quote from an average-Joe in the US who said ‘I pay $15/month for internet. That’s pocket change! I don’t even think about it”. Indian middle class needs to feel this way about using the internet (we’ve already reached this in how we view cable TV, which has penetrated 60 million households). Then, we’ll be in a mode where people discover the power of the net, as opposed to using it in a limited way for point-applications (email, sporadic travel bookings and job searches).
There is no shortage of compelling content/applications (travel, stock-trading, jobs, shopping sites have been around), though regional language content can still be strengthened. In my view, access is the main bottleneck. Drawing lessons from two notable Indian success stories – cable TV and cellphones – here’s what I’d like to see over the next 1-2 years in the internet access space:
Far higher competition: Currently cable-operators and PSU telcos dominate the last mile. Frankly, both of these suck big-time! The former are controlled by political parties and local goons. The latter’s crappy service is a turn-off. We need the equivalent of Bharti and Reliance – national players with deep pockets and a sharp focus on growing the market. These players are so busy with the booming cellphone business that they haven’t devoted adequate focus to growing their data services. I don’t blame them, but am hoping this will change once the government forces sharing of last-mile copper by PSUs and wireless profit growth slows. I am all for violent competition between cable TV, satellite TV, fixed line incumbents, ISPs and wireless operators.
Multiple access technologies: Think of what CDMA did to GSM (ignore the regulatory back-door entry part). I am bullish on Wimax, as that can break the last-mile gridlock. However, I expect initial versions of Wimax to be too expensive to meet Indian price points. I look forward to innovative companies that can come up with desi Wimax technologies and business models.
Shift in mindset from PC-price to entry-cost & monthly charges: There’s been a lot of noise on the Rs. 10,000 PC. That’s a good thing, but the more realistic price for a PC with decent functionality remains over Rs. 20,000. Simple sum-of-parts math (think monitor, windows) will make sure PC price doesn’t fall fast enough. Alternate configurations & access devices appear too niche and hard-to-use. Instead, what we need is a mindset and marketing change, to ‘pay Rs. 1000 unfront, and Rs. 500/month for PC + internet’. Loans and EMI-payments for PCs exist even today, but aren’t yet packaged and marketed the way wireless companies say ‘lifetime prepaid phone for Rs. 999’. I realize that the math still needs to add up, but people will be creative if the market is large enough.

PCs in education: Middle class parents will buy a PC, if they think their kid’s life prospects will improve because of this. This hasn’t gained scale yet, as most schools don’t have the notion PCs and internet being integral to learning and class projects. Computer training gains focus only in college, that too in the parallel universe of training institutes. Maybe, the industry body and government agencies can play a role in educating our educators. Once the PC enters the home (using education as an excuse), both kids and parents will explore and discover new applications (including some not-so-clean ones!). Once home-internet-access is fixed, usage should follow. Our growth trajectory will still be different from the West, due to low cost of intermediation (cheap labour) and a lack of a self-service culture (where else do you have attendants manning vending machines!). But, who knows! Here’s a question I’d like your views on. Who will be the first Indian internet company to reach $100 million in revenue? And please don’t tell me ‘Indian Railways’!

No comments: