Originally Published 2011-07-04 00:00:00 Published on Jul 04, 2011
Telecom companies are stuck with falling ARPUs. Indian researchers from two unlikely companies show them new tricks of the trade
When in doubt don't write, just speak
Using Bluetooth Ram Khelawan Pasi spirits out the Salman Khan blockbuster of a song Munni Badnam Hui from his friend Rajkumar Lodhi’s phone. Pasi has over 300 songs in his phone and listens to it while he works. For city slickers it’s an image that’s not in one bit unusual. But 43-year-old Pasi is a farmer, cannot read or write and has never left his village of Maswasi in the Sikandarpur Karan block in Unnao district of Uttar Pradesh. For that matter, neither did his father or grandfather. Pasi and his wife work on their tiny piece of land growing vegetables and occasionally cereals and grains. On a lucky day he earns Rs70. On other days he considers himself lucky if he earns Rs50. His two children are the first ones in recent generational memory to ever attend a school.

An illiterate farmer earning less than Rs100 a day having an uncommon expertise in operating a Bluetooth-enabled phone is not a typical use case that telecom operators, MVAS providers and business development professionals would include in their strategy presentations on increasing the rapidly falling ARPUs of telecom companies. For Pasi and countless others like him to have mastered the relatively complex technology of a mobile phone to such an extent as to independently engage in peer-to-peer transaction is not only a fundamental sociological insight but also a robust, scalable and profitable social business model waiting to be explored.

Yet for telecom companies and MVAS providers Pasi is a customer they wish never existed. He is a pre-paid subscriber, like over 95% of the Indian mobile phone owners, with a minimum balance of Rs10-25. Not enough to consume their VAS packs at price-points ranging from Rs30-50. Not that the VAS packs with their offerings of cricket and stock market alerts, astrological services and animated wallpapers of scantily clad heroines and CRBTs and RBTs of popular Bollywood and Bhojpuri songs interest Pasi enough to pay for it.

Pasi is clear what he doesn’t want from his mobile phone. Songs, videos and wallpapers, he says, are always available ’for free’ from friends, hinting at the Wild Wild West nature of the digital landscape of the Indian boondocks. It’s something that will make the suits betting their money on Digital Rights Management (DRM) and Intellectual Property Rights (IPR) sweaty browed and hot under the collar. He is also clear what he wants - something that will earn him more money or at least help him save some. He is willing to pay anyone who tells him over the phone the latest rates in the mandi, help him find out why his crops are getting infected, the best price his milk will fetch or the exact amount of food-grains and commodities dispatched to his neighbourhood ration shop so that the kotedar doesn’t cheat him.

Pasi is essentially saying two things. Give me information that I can use and give it to me in an audio format. There are 350 millions Indians like Pasi - illiterate marginal farmers, underemployed, but increasingly connected through mobile phones and willing to pay for information that is useful to them.  One would have imagined the fleet-footed telecom and MVAS companies falling over themselves trying to cash in on the opportunity. But they have been obsessing far too long over the formulaic ABC (Astrology, Bollywood and Cricket) content offering. So they haven’t.

But two unlikely companies have. Giants in their own right, nobody expected IBM and TCS, companies known for their enterprise solutions and high quality customised software packages, to cross the rubicon of the Spoken Web. All the more commendable considering that in both cases it was an all-Indian team that cracked the puzzle. This author met some members of both the teams at the recently concluded international conference of digital professionals and researchers in Hyderabad. Digital evangelists have long been talking about how the path to breaking down the literacy and language barrier rested on the ability to move away from a text-based World Wide Web dominated only by information exchanges to one driven by spoken forms dominated by transactional exchanges.

The five-member team from IBM Research Labs - Sheetal K Agarwal, Dipan Chakraborty, Arun Kumar, Amit Nanavati and Nitendra Rajput - evolved a completely new protocol called Hyperspeech Transfer Protocol (HSTP), similar to the Hypertext Transfer Protocol (HTTP) that allows for the seamless browsing of the web, to enable users to browse through voice applications and transact across two or more of such applications. For those like me obsessed with the nuts and bolts of the HSTP, VoiceXML and hypermedia architecture the full paper can be accessed at

The soft-spoken Manish Gupta, head of the IBM Research Labs, is an IBM veteran. He joined the Big Blue in 1992 in their New York labs and has never looked elsewhere. "Look at the scale of India. Compare the Internet penetration levels and the mobile penetrations levels. Innovation is bound to happen here first and adopted by the world later," he says confidently. The proof of the pudding, however, lies in eating it. The team took that adage to heart and did two pilot projects, one in 20 villages of Gujarat called Avaaj Otlo, and another in six villages of Andhra Pradesh.

The principle is simple, yet it’s a paradigm shift. Every single mobile phone number is unique. So the team figured out that every single number is as good as a Unique Resource Locator (URL), the virtual billboard that helps you navigate correctly to your favourite website. And so was born the VoiceSite. Like a typical WWW environment you need application generators, which the team calls VoiGen, and hosting services, called VoiHost, which can reside within the current telecom network. So a user calls up a number, much like an Interactive Voice Response System (IVRS), and telephonically logs on to the VoiceSite. Once inside, through a voice navigation system, the user can contribute local content, navigate available content - in much the same manner as a surfer would browse through web content - place audio advertisements or listen to them, pause, rewind, re-play and close the application whenever s/he wants it.

The response was phenomenal. Frustrated by persistent pest attacks on his cotton crop, farmer Babu logged on to the VoiceSite asking someone for a solution. In less than three hours another ’senior’ farmer from a different village responded to Babu’s problem and gave him the solution. The VoiceSite is actually a bustling cackle of opportunities - cries of a newly born grandson being recorded so that his grandmother can hear him, a plumber advertising his skills by posting a voice resume and, of course, random people singing random songs.

"The most interesting use of the VoiceSite that we came across was from a man who posted his matrimonial advertisement," laughs researcher Amit Nanavati. "The best part was that within a couple of hours someone else had posted a response saying that this man is known in his village to try and trick young girls into marrying him." The two projects were in operation for eight months. Over 8,000 people made 110,000 calls.

Bespectacled and professorial Dr. Arun Pande is the head of the TCS Innovations Lab. His team has also tapped into the potential of the Voice Web, but has focused it primarily on farmers and integrated text as well visual components on to it. Called mKrishi the voice platform combines several technologies, ties up with local and national resource persons and institutions to bring information in local languages about weather, fertiliser requirements, steps to control pests, grain prices in local markers to the farmer’s low-end mobile handsets.

"It allows farmers to send queries in their local languages, as well as images and voice activated SMS through a mobile phone and provides personal responses with advice or relevant information in these languages," says Dr. Pande. The pilot project, implemented in the Raigad district of Maharashtra, has seen such encouraging response that Dr. Pande and his team have now customised the mKrishi platform to cater to the basic health needs of a rural population.

"During interactions with small and marginal farmers, we realised that they had to travel large distances to receive treatment. We tuned mKRISHI platform for delivering primary health care to villagers. Instead of farmer and his soil analysis details, we provided a patient’s health record and answers to his queries. The doctor remotely prescribes medicines to the patient through village health workers," explains Dr. Pande. "I have had a farmer sending me a video of his ailing crop and continue it to show the rash that he was suffering from in his heel. The manner in which a poorly educated farmer took to this technology astounded us."

Pasi doesn’t know the Big Blue or the Tata giant. Nor has he heard about Dr. Pande or Dr. Gupta or the numerous scientists working to make his wish a reality. But one day he will. And on that day it’s anybody’s guess whether he will write a thank you note or say it over the phone and voicecast it to the rest of the world. Till then let’s keep our fingers crossed.

(R. Swaminathan is a National Internet Exchange of India (NIXI) Fellow.)

Courtesy: Governance Now
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.