Mohit Bhatnagar, Airtel: His 646- service broke the record for calls logged on election results day

May 13, 2004: It's election results day. The initial results indicate a wild deviation from what the exit polls suggested. The uncertainty has bred tension on the streets as well as in the boardrooms of Corporate India. But Bharti TeleVenture CEO Sunil Mittal is an odd person in that glum milieu. That's because he has a big reason to smile - his company has created a record of sorts. About 130,000 people have called Airtel Live's 646 voice service to check out the election results. It's the highest number of calls in a day that any operator has fielded so far on such a service.

The service's success hasn't been a one-day phenomenon. "Voice portal minutes have grown 60 per cent over the last quarter," says Mohit Bhatnagar, vice-president (new product development and alliances), Airtel. Even Hutch (123-service), Airtel's closest competitor, claims that its two-year-old voice service is picking up momentum. Others, like Idea Cellular (456-service) and BPL Mobile (Just Call service), who have just started their own voice services, say they too are banking on expanding revenues from value added services (VAS).

Voice services, a part of VAS, are designated numbers consumers can call to request for information or downloads. The service is based on speech recognition technology. The other VAS offerings are data-based. That is, they are available through SMS on GSM (global system for mobile communications) phones or applications accessible through GPRS (general packet radio service, a faster GSM standard). So far, Reliance Infocomm is the only operator that is not banking on voice services. Instead, its CDMA (code division multiple access, another system) service is putting all emphasis on data-based VAS. (See 'You Can Also Talk', BW, 5 July 2004)

In the coming year, GSM operators are likely to focus on these voice services. Already, Hutch has raised the pitch of its advertising campaign for its voice services (featuring cricketer Rahul Dravid). So has Airtel.

There are good reasons why voice services are critical to GSM players. For one, operators are earning a pretty penny on them. While VAS as a whole earned 3.4 per cent of the industry's topline in 2003 (says data and analysis provider Gartner), it brought in about 10-15 per cent of the bottomline.

And voice services now account for about half the VAS revenues (excluding those from peer-to-peer SMS), say industry sources.

Almost all the telecom operators have reported healthy trial rates. "About 40-50 per cent of mobile users have already tried it once and repeat usage is also high," says Arvind Rao, chairman and CEO of OnMobile, the company that created the technology for all these voice services. That's understandable, especially since most Indians still find it easier to call in rather than download data. Operators say most users find it difficult to remember SMS codes, and the number of GPRS users in India as a share of the total subscriber base is low.

There's another important reason why telecom operators will look to drive voice services. Although the overall market is growing, the expansion is increasingly bringing in subscribers with lower disposable incomes.

Gartner's telecom analyst Kobita Desai says: "Operators can expect their average revenue per user (ARPU) to be under consistent pressure. While mobile applications may not have a significant impact in yielding incremental ARPU, they would help in stemming the southward drift of ARPUs, or at least marginalise it."

But here's the big question: are Indian telecom players getting carried away by the allure of a short-term opportunity? Are they ignoring the perils of too much voice and too little data?

Voicing Concerns

The reason operators need to heed the questions is that once volumes pick up, the economics of voice services will change. To figure out how, we need to understand the cost structure. The biggest upfront cost in starting VAS is the cost of the platform itself. The content for most individual services like cricket scores on SMS are on revenue share deals and hence don't attract any upfront cost. But one needs to invest in an SMS centre to provide SMS or a multi-media messaging service centre for MMS. But OnMobile has built the voice portals for all the leading players on a revenue-share basis. "So the capital expenditure, stretching across numerous services, will need to be offset by total volumes on the platform, and not so much by individual services," says Balu Nayar, associate vice-president and head of VAS, Hutch.

Hutch and Airtel have been quick to seize the opportunity — their new campaigns are focussed on voice services

The problem is that voice services take up a lot of network space and that can put strains on profitability. Operators claim that none of the voice services are choking up their networks yet. But as volumes move up, this will change.

According to consulting firm Analys, a minute of voice telephony could take up over 14 times the network capacity of an MMS, and over 900 times that of an SMS. Revenue per kilobit is a metric that telecom operators will not be able to ignore soon, says Desai.

As voice volumes increase, GSM operators can either increase network capacity or shift users to data-based services. The first would come at a substantial cost and hence directly affect an operator's bottomline. The better way to address the problem would be to shift to data, the alternative that uses network capacity more efficiently. Also, more services - like email, MMS - can be offered through a data portal.

For the moment though, it won't be easy to push data services. The limited number of data-enabled (GPRS) handsets is a significant handicap for GSM operators. Though there are 1.25 million GPRS handsets in India, only about 300,000 have signed up for the data services that run on that technology. And herein lies a paradox. Operators cite the low user base as the main reason to go for applications available on voice and SMS, and not on GPRS portals. On the other hand, users are deterred from taking to GPRS services because of the steep price tag they come with at present. So though 30 per cent of the new handsets sold in India are GPRS-enabled, not many are taking to the service. If usage has to grow, the pricing has to be sorted first.

Pricing It Right

Arvind Rao, OnMobile: The voice behind all GSM voice applications

One would expect the pricing of VAS at this early stage to be attractive enough to develop the market. Yet, on the face of it, GSM operators seem to be skimming the market - on both voice and data services. BPL Mobile offered exam results on SMS at Rs 7 a pop for a few months, and then increased the price to Rs 10. Idea Cellular has priced its voice portal services at Rs 4.99 a minute. Now compare these to the Rs 1.50 or less one pays for a peer-to-peer SMS and you get a sense of the pricing.

Desai says Indian operators could well be committing the mistake that Korean operators made in their initial years. While trying to grow ARPU, the largest Korean operator, SKT Telecom, had priced its entire basket of VAS on the higher side. The result: the anticipated jump in ARPUs did not materialise, as consumers did not increase the mobile services' share of their wallet.

There is more data to support the claim that Indian operators are getting carried away on the matter of pricing VAS. Consider this: Indian GPRS services are priced at Rs 99 and Rs 499 for limited and full-fledged access - a shade higher than the average national ARPU of Rs 451. US operator Nextel Communications has an average ARPU of $70 and its email product is priced at $5, or 7 per cent of the ARPU. Another US operator Verizon's ARPU is $50 and its email product is priced at $5. Japanese operator KDDI offers a basket of VAS, including email and photomail, at a flat fee of $11, 15 per cent of its monthly ARPU.

Indian operators say their target market can afford these services. It's true that GPRS services are aimed at the premium segment, and not the mass market. Besides, the content bought by the operator often comes at significant costs - like those for Spiderman games or cricket World Cup rights - and they have to be recovered.

Hutch has its own rationale for premium pricing. When it introduced Java games, it had a seven-month lead over the rest of the field. So it priced games at Rs 50 each, which was almost as high as the monthly rental for some other services on the network. To induce trials, Hutch also offered a limited-play version of each game for Rs 20 (a proposition akin to a shampoo sachet's in the FMCG space). In the first quarter after introduction, a third of the users who played the games opted for the limited-play version. But by the next quarter, that shrunk to a tenth of the users. Soon afterwards, it dipped below that.

Says Nayar of Hutch: "Our consumers found far greater value in an unlimited bottle of shampoo than the low-priced sachet. In the games category, the price benchmarks our consumer looks at are not related to the prices of other services, but to other avenues of entertainment."

For now, CEOs like Mittal are relishing the first flush of success. But there is a call waiting for them. Sooner or later, they will have to pick it up.

Inputs by Irshad Daftari.