Getting Online: The Telecom Revolution and the Arrival of Search, 2000 to 2010

A New License to Compete

The previous era in this series, the opening that liberalization and satellite television forced from 1991 to 2000, left India's telecom sector with a working mobile network but a licensing regime that was close to strangling it. Operators who had bid aggressively for a handful of cellular licenses in the mid-1990s owed the government large, fixed annual fees regardless of how many subscribers they actually signed up, a structure that made a low-priced mass mobile market close to impossible and left several licensees close to bankruptcy by decade's end.

The New Telecom Policy of 1999 addressed that directly. Announced by the Vajpayee government and effective from 1 April 1999, it let operators migrate from a fixed license fee to a revenue-share arrangement, paying the government a percentage of what they actually earned instead of a sum fixed years in advance. The migration package for the eight existing cellular licensees, approved that year and effective from 1 August 1999, set the license fee at roughly 15 percent of adjusted gross revenue, in exchange for giving up the guarantee of a limited number of competitors in each circle.

A second structural change followed in November 2003, when the Department of Telecommunications, on the Telecom Regulatory Authority of India's recommendation, introduced the Unified Access Service License. The new license was technology-neutral and let a single operator offer fixed-line and mobile service together across any of the country's 23 defined service areas, removing the old separation between basic and cellular licenses and opening the door to the multi-operator, multi-technology price competition that would define the rest of the decade.

A parallel legal foundation was laid the same year the policy reset began. The Information Technology Act, 2000, received presidential assent on 9 May 2000 and came into force on 17 October 2000, giving electronic records and digital signatures legal recognition for the first time and making India the twelfth country in the world with dedicated legislation of its kind. Without it, the online railway bookings, the digital marketplaces, and eventually the contracts this decade's businesses would rely on had no settled legal ground to stand on.

Reliance Rewrites the Price of a Phone Call

Reliance Infocomm gave the new licensing regime its first real test. The company commissioned the fiber backbone for its CDMA network on 28 December 2002, and launched nationwide CDMA2000 mobile service in the months that followed, entering a market that GSM operators like Bharti Airtel, Hutch, and the state-run BSNL had until then largely had to themselves.

The offer that made the entry impossible to ignore was the Monsoon Hungama scheme, launched in mid-2003: a mobile handset for Rs 501 (about $11 at the time) that came bundled with Rs 100 (about $2) of free talk time, at a moment when a basic mobile connection had generally cost several times that to acquire. Within nine months of the May 2003 CDMA2000 launch, Reliance had captured close to 20 percent of the Indian mobile market, a pace of share gain that forced every GSM incumbent into matching price cuts of its own.

The subscriber numbers that followed were the clearest evidence of what the combined effect, cheaper licenses feeding cheaper handsets feeding cheaper tariffs, had done. The Telecom Regulatory Authority of India's own quarterly reporting put cellular and fixed-wireless subscribers at roughly 28 million in early 2004; by the end of 2008 the combined GSM and CDMA subscriber base stood at about 375 million, more than a thirteen-fold increase in under five years. For most first-time owners, that handset, not a desktop or laptop computer, was the first personal computing device they had ever held.

That new mass of connections also produced a distinctly Indian habit: the missed call. Because most subscribers were on prepaid plans with limited talk time, dialing a number and hanging up before it connected became a free way to send a signal, call me back, I have arrived, I am thinking of you, without spending a rupee. By 2008 one estimate found more than half of Indian mobile users placing missed calls habitually, and the practice was reportedly occupying more than 30 percent of network capacity while generating no revenue for operators at all. Operators' own mobile VAS, ringtones, caller tunes, cricket-score and devotional SMS alerts, became the first real mobile advertising and content market in the country, built on the same prepaid, low-balance behavior that made the missed call so useful.

The Tariff War That Closed Out the Decade

Years of UASL-driven competition among ten or more operators per circle had already pushed voice tariffs down substantially by 2009, but the convention of billing every call to the nearest full minute, so a five-second call cost the same as a fifty-five-second one, had survived. Tata DoCoMo broke it when it launched service in Chennai on 24 June 2009 with per-second billing at one paisa a second, a pricing structure competitors had not offered before.

The response was immediate and industry-wide. Rival operators matched and then undercut the new billing model, with some rates falling as low as half a paisa per second for in-network calls by December 2009. The effect on adoption was sharp: mobile penetration is reported to have surged from 29 percent to 43 percent of the population within a single year, and the industry was adding roughly 15 million new subscribers a month, at the time the fastest net-addition pace of any mobile market in the world.

Television did not lose ground to any of this; it grew alongside it. The government had permitted direct-to-home satellite reception in November 2000, and Dish TV, part of the Zee group, launched India's first DTH service on 2 October 2003, deliberately targeting towns and villages that cable operators had not reached rather than competing head-on with them in the metros; it had signed up 350,000 subscribers within two years. Prasar Bharati, the public broadcaster, added its own free option, DD Free Dish, in December 2004, extending Doordarshan's reach and that of a growing list of free-to-air private channels into homes that owned a dish but paid no monthly cable bill.

Cybercafes remained, for a large part of the decade, the only place most first-time internet users could actually get online, since phones of this era were voice-and-SMS devices rather than data devices, and home broadband stayed a minority experience even as mobile ownership multiplied. Sify's iWay chain, the country's first nationwide branded cybercafe network, grew from under 1,000 outlets in mid-2003 to more than 3,100 across 149 cities, peaking around 3,300 cafes in over 150 cities by 2007, though industry-wide cybercafe growth had actually peaked earlier, around 2002, and slowed for the rest of the decade as stricter identity-logging rules added friction for operators. India's internet-user base grew from an estimated 0.8 million in 2000 to somewhere in the range of 60 million by 2010, a wide range across different counting methods, but still a small fraction of the population next to the hundreds of millions who by then owned a phone.

Google Comes to Bangalore

Google opened an engineering office in Bangalore in 2004, its first full research and development center outside the United States, starting with roughly five employees and an initial investment in the low single-digit millions of dollars, tasked with data mining, data warehousing, and business-intelligence work alongside India-specific products. It was a small bet at the time, dwarfed by the scale of the IT and BPO campuses already operating nearby.

What made the bet pay off was less the office itself than the timing. Mobile growth was putting a first personal device into the hands of tens of millions of Indians every year, and even though most of those phones could not yet browse the open web, the smaller but fast-growing population reaching the internet through a cybercafe or a slow home connection found Google's stripped-down, ad-light homepage loaded acceptably even over the narrow-band links most Indian users still had.

The scale of the result was measured directly a few years later. In June 2008, per comScore's tracking, Indians ran 1.242 billion searches in a single month, and Google Sites accounted for 81.4 percent of them; Yahoo, the next largest, held just 9.4 percent, and the strongest homegrown portal, Rediff.com, captured only 1.5 percent. Despite holding more than 15 percent of the world's population, India accounted for under 2 percent of global search volume at that point, a reminder of how early the country still was in its internet build-out even as the contest for search leadership within that smaller pool was already decided.

That dominance sat in direct contrast to the destination-portal model that had defined Indian internet use through the previous decade. Rediff.com, launched in 1996 by Ajit Balakrishnan as India's first major homegrown web portal, had built its business on being a single destination for email, news, and shopping, and had listed on NASDAQ in 2000 at a peak valuation above $600 million; Indiatimes followed a similar all-in-one model. As search became the default starting point for almost any online task, the portals' destination model inevitably lost ground to a general-purpose lookup tool, even where, as with Rediff, the businesses themselves survived.

Discovery Before the Search Box

Google's dominance was general-purpose search; several narrower, purpose-built Indian ventures had already solved specific discovery problems years earlier, and they stayed relevant precisely because a generic search box did not do their job as well. Finding a job, a spouse, or a nearby shop needed a structured listing, not a ranked page of web results.

Naukri.com, founded by Sanjeev Bikhchandani and launched on 27 March 1997 as a division of Info Edge with a seed capital of Rs 2,000 (about $55 at the time), digitized the job classified, going commercial that October with a database pulled initially from print listings; it became India's leading job portal years before most of its users had ever run a search on Google.

Murugavel Janakiraman built the matrimony equivalent almost by accident. In 1997, while working in the United States, he added a matrimonial section to a Tamil-community portal, SysIndia.com, as a side project; when it drew more traffic than anything else on the site, and after it matched him with his own future wife, he spun it into a dedicated business, BharatMatrimony, in 2001, digitizing a discovery task that Indian classifieds and family networks had handled for generations.

Justdial pushed the same instinct further into the offline world. V.S.S. Mani started the business in 1996 with Rs 50,000 (about $1,400) as a telephone-based local search service: callers dialed a number and asked a human operator for a business listing rather than typing a query. Justdial added a website in 2000 and an internet-plus-mobile version in 2007, but the phone-in model mattered on its own terms because it worked for the much larger share of Indians who were not online at all, extending local business discovery to a customer base search engines could not yet reach.

Booking a Seat, Buying a Book

Indian Railways gave the country its first taste of transacting online at national scale. IRCTC, the railway's dedicated e-commerce arm established in September 1999, launched online ticket booking on 3 August 2002. The service processed just 27 tickets on its first day, delivering the physical ticket by courier since e-tickets did not yet exist, but it proved that a government service, not only a startup, could move a paper-bound Indian institution onto the web, and online rail booking grew into one of the most-used transactions in the country over the years that followed.

Retail e-commerce arrived more haltingly, and from outside India first. Baazee.com, an auction-style online marketplace, had built roughly a million registered users by the time eBay bought it outright on 2 August 2004 for about $50 million, at a moment when India had only around 17 million people online and was projected, and turned out to be roughly right, to reach about 30 million by 2006. The acquisition made eBay the country's earliest major e-commerce operator, well ahead of the names that would come to define the category.

The company that eventually eclipsed eBay India started at the opposite end of the decade and the opposite end of ambition. Sachin Bansal and Binny Bansal, two engineers who had been working at Amazon in Bangalore, founded Flipkart on 22 October 2007 with about $4,200 pooled from each of their families, roughly Rs 400,000 in total, and opened as an online bookstore run out of a two-bedroom apartment in Koramangala. Within about a year the site was processing roughly 100 orders a day, a small number against the country's population, but early proof that Indians would pay online for a physical good delivered to their door.

What connects IRCTC, Baazee, and Flipkart across the decade is less their individual scale, all three began small even by the standards of their own time, than the sequence they establish. A state agency proved the payment-and-delivery loop could work at all; an acquired marketplace proved a foreign buyer would pay real money for an Indian online audience; and a two-person startup proved the model could be built from nothing. Each step lowered the perceived risk for the next entrant, a pattern that would repeat at far larger scale once smartphones and mobile data arrived in the following era.

The Boom Behind the Screens

None of the subscriber and search numbers above happened in an economic vacuum. Outside the sharp dip tied to the 2008 global financial crisis, when growth slowed to about 3.1 percent, India's real GDP grew close to or above 9 percent in 2005 and 2006 and stayed above 7 percent for most of the years in between, the strongest sustained stretch the economy had recorded since the 1991 reforms opened it up. That growth put disposable income into the hands of exactly the first-time buyers Reliance's Rs 501 handset, and the tariff wars that followed it, were built to reach.

A large share of that growth traced back to one export industry. NASSCOM figures show the sector's software and services exports crossing $12 billion in fiscal 2004-05, up 34 percent on the year before, and the broader IT-BPO industry, exports plus the domestic market, reaching roughly $52 billion in revenue by fiscal 2008, a 28 percent increase on the prior year, with software and services exports alone crossing $40 billion. Bangalore, Hyderabad, Pune, and the National Capital Region built entire employment bases, and a consumer class with disposable income and English-language office jobs, on the back of that export engine.

The two threads met directly in the workforce every dot-com and BPO firm hired from. A software or BPO job was, for many urban Indians, the first source of income tied directly to a computer, a headset, and an internet connection rather than to a television set or a newspaper. For that population, and increasingly for the far larger population reached only by an inexpensive mobile connection, the phone rather than the personal computer had become, by the end of the decade, the device most Indians actually owned. The first, small Indian digital-advertising agencies and mobile-VAS aggregators built their businesses selling portal banner space and operator-billed subscriptions in this same window, though the category remained a modest fraction of a still overwhelmingly television-and-print advertising economy.

The License Regime's Shadow Side

The same administrative, non-auction method of allocating scarce telecom licenses that had made the UASL regime cheap enough for Reliance, and a wave of operators after it, to compete on price through the decade returned in a far more consequential form at the era's very end. In January 2008 the Department of Telecommunications issued 122 new 2G mobile licenses to a dozen companies, pricing the licenses and associated spectrum at rates fixed back in 2001, years before the subscriber base had grown into the hundreds of millions, and doing so without the competitive auction that TRAI, the Law Ministry, and the Finance Ministry had each, in different ways, recommended. Subsequent scrutiny found that 85 of those 122 licenses had gone to entities that did not meet the department's own eligibility conditions.

The consequences played out mostly after this era closes, but the audit that defined the controversy landed right at its edge. In November 2010 the Comptroller and Auditor General of India released a report estimating that the underpriced allocation had cost the exchequer a notional loss of about Rs 1.76 lakh crore, roughly $39 billion at the time. The figure, and the licensing process the telecom minister of the day had overseen, became the defining political scandal of the following Parliament and occupied investigators and courts for years afterward.

For the purposes of this decade's story, the point is narrower than the scandal itself. The same permissive, license-first policy toward telecom access, the one that priced Reliance's CDMA entry cheaply enough to fund a Rs 501 handset, that let a dozen operators fight a tariff war down to half a paisa a second, and that turned the mobile phone into the personal computer of a nation within a decade, carried a real cost when the same discretion was applied to a genuinely scarce resource just as its market value was rising fastest. The cheap access that built this era's marketing and discovery surfaces was itself a policy choice, made, and in this instance mismanaged, by people, not a natural feature of the market. What the license regime had built by 2010, a mobile-first population and a Google-first search habit, is the base the smartphone and the Jio shock would build on next.