From Oil to Jio

Part II

For Part I on Reliance’s origins click here.

My co-conspirator Anmol and I bantered on WhatsApp while we watched the Reliance AGM early in the morning on the 15th: "You know if you read the Jio logo backwards, it reads oil." I made the connection only in half-jest. Mukesh Ambani specifically called out the link when Jio was launched in 2016, "For Reliance... data is the new oil".

The story of Reliance during Dhirubhai Ambani's lifetime was one of backward integration, starting with polyester yarn and slowly securing the supply chain expanding first into petrochemicals, then petroleum refining and finally oil and gas exploration. Following Dhirubhai's death in 2002, Reliance had to find its footing again, especially in light of the company's split, a few years later, following a bitter feud between brothers Mukesh and Anil. Where would the new Reliance(s) look to secure their future? Telecom has been central to this story, which on the one hand feels like a long saga, but on the other, is clearly still unfolding.

Expanding to telecom

The story of Reliance's foray into telecom, unsurprisingly, starts with liberalization. In particular, it was under Atal Bihari Vajpayee's government that the telecom sector was significantly deregulated and private and foreign ownership limits significantly amended. It was during this period that Reliance first acquired telecom licenses. The young, technocratic BJP leader, Pramod Mahajan, who was telecommunications minister between 2001 and 2003, played a major role in implementing the 1999 Telecom Policy, which really changed India's telecom landscape.

The massive change the policy led to is immediately visible from the Telecom Regulatory Authority of India's data below. It was a classic example of how unclogging India's regulatory cholesterol led to massive growth, since there was underlying pent-up demand. Mobile phones became a symbol of leap-frogging; instead of waiting to get a landline (famously, during the peak years of the License Raj this wait could last for years), the country just started using mobile phones. By the end of 2019, India had 1.15B mobile subscribers.

This article in The Caravan (which on the whole I found really problematic) is interesting in the little snippets it provides about the initial thinking on telecom within Reliance, in particular, by Mukesh Ambani. Initially unconvinced and unenthusiastic about the sector and the company's ability to build a profitable business in it, he slowly really came around. When I was researching Reliance, I ended up watching this old Simi Garewal interview with Mukesh Ambani dating back to ~1997 (it's such a throwback to a simpler media era!) where he specifically calls out telecom as an area of focus as Reliance looks to the future. Ambani and Mahajan were, by many accounts, close and when Reliance Infocomm was launched in 2002 soon after Dhirubhai Ambani's death, it was Mahajan who was the guest of honour. But the company's telecom arm had a turbulent start, and not because the all too familiar allegations of favouritism, and crony capitalism dogged Reliance's entry into the sector.

A very public cold war had broken out between the two brothers not long after Infocomm’s launch, with control of the telecom arm, in particular, becoming a sticking point (Anil had pointedly not attended the launch). A lot of ink has been spilled on the details of the sibling rivalry, which I won't recount except to say that in 2005 the core company was split into two. Mukesh Ambani got the petrochemicals business and got to keep the Reliance name, Reliance Industries Ltd (RIL) ; Anil took control of the telecom arm, along with a bunch of other entities, and called his company the Reliance-Anil Dhirubhai Ambani Group (R-Adag). The two also signed a ten year non-compete, which should have technically prevented further antagonism. But that was not to be. By 2008, the telecom arm under R-Adag had become the country's second largest cellular network and was planning to acquire MTN, the South African telecom giant to expand its footprint. RIL threw a spanner in the works by claiming it had right of first refusal. Eventually the acquisition was called off and in 2010 the two firms scrapped their non-compete agreement, giving RIL leeway to re-enter the telecom sector.

Betting on data

The RIL and R-Adag non-compete was scrapped just as the Indian government began auctioning 3G and broadband wireless-access (BWA) spectrum, used primarily for high speed data access. Most of the incumbents focused their energy and money on the 3G spectrum for voice, text and data. The BWA auction was the first of its kind and attracted as many as eleven bidders, including the major telecom players at the time (including R-Adag). However, the biggest winner of the BWA auction was a pretty unknown company called Infotel, which won licenses for all the 22 regions the government was auctioning off. None of the 3G spectrum owners had a national footprint, which gave Infotel a massive advantage. Infotel at the time had annual earnings of only some ~$32K and a single internet subscriber, but did have the requisite bank guarantees to bid for the spectrum. "Who are these Infotel guys?" someone asked here when the auction results were announced. In a couple of days, the answer was clear. RIL bought a 95% stake in the company for ~$1B and made it a subsidiary to re-enter the telecom business.

A lot of journalistic and competitor hackles are now raised about this move when people now talk about Jio, including arguments that this fronting for RIL led to a loss for the exchequer. I have a massive dislike for the "notional loss" way of thinking that the Comptroller and Auditor General of India championed under the UPA government (much to the BJP's benefit) because it's a very hindsight 20:20 way of looking of things. Similar arguments are now applied to BWA auction. One competitor said, "We thought it was an ineffective, useless, underfunded company," arguing that had his company known who it actually was they would have bid more, which, in turn, would have increased the amount of money that the government collected. In April 2012, TRAI recommended letting companies with internet-only licenses upgrade to unified-access licenses. RIL immediately took advantage of this, paying migration and entry fees, but faced charges of backdoor entry (just as it had done when it first entered telecom). I find this line of reasoning pretty asinine, even though there's a case to be made about RIL initially not operating within the spirit of the rules. It is also worth remembering that the government actually ended up making 3x what it had originally projected from the 3G and BW spectrum sale.

It's also easy, post the success of Jio, to state that RIL bent the rules in its favour, but this completely understates the risk RIL was taking by basically attempting to leapfrog existing telecom technology. RIL's competitors were still rolling out the 3G network. The Wall Street Journal noted at the time, "By the time Mukesh Ambani builds a 4G wireless business, rivals will have had the chance to sign up millions of customers for 3G services, leaving a smaller pool of potential broadband subscribers. Also, 4G technologies are still being fine-tuned, whereas 3G networks have been up and running for years in other parts of the world. And 4G devices will likely be more costly than 3G ones initially, because there will be a smaller universe of manufacturers."

In hindsight, though, this turned out to be a massive advantage for RIL. (R-Adag in contrast, for example, was saddled with CDMA technology, and had massive capex spend to migrate its customers to the more popular GSM and LTE networks, a major reason why Anil Ambani's group is now essentially shuttered. Mukesh Ambani probably dodged a bullet by not getting that half on the company!) Ben Thomspon laid out the benefits of leapfrogging really well in his write-up on Jio:

"The key to understanding Ambani’s bet is that while all of the incumbent mobile operators in India were, like mobile operators around the world, companies built on voice calls that layered on data, Jio was built to be a data network — specifically 4G — from the beginning

• 4G, unlike 2G and 3G, does not support traditional circuit-switched telephony services; voice calls are instead handled the same as any other data.

• Because everything is data, 4G networks can be built with commodity hardware in a way that 2G and 3G networks cannot.

• Because Jio was offering a data network, voice calls, which are relatively low bandwidth, were the cheapest services to offer, and capacity was effectively infinite.

To put it another way, Jio was a bet on zero marginal costs — or, at a minimum, drastically lower marginal costs than its competitors. This meant that the optimal strategy was — you know what is coming! — to spend a massive amount of money up front and then seek to serve the greatest number of consumers in order to get maximum leverage on that up-front investment."

Thompson frames this as a classic Silicon Valley bet, which it is, but it is also a very Indian, and very Reliance, move. Arun Sukumar, whose book Midnight's Machines is a political history of technology in India, pointed out to me that Ambani's bet on Jio follows the philosophy of another famous Gujarati, Vikram Sarabhai, who liberally applied the terminology of leapfrogging to the Indian context. "Our national goals involve leapfrogging from a state of economic backwardness and social disabilities attempting to achieve in a few decades a change which has incidentally taken centuries in other countries and in other lands," Sarabhai said. Sarabhai, was of course, talking in the context of India's state-led and rarefied space program. Ambani's leapfrogging has been a private venture that has had a more immediate impact on the life of every day India, but is no less audacious.

When Mukesh Ambani had spearheaded Reliance's initial foray into telecom before the company split, his vision had been to make the cost of a phone call as cheap as a postcard. When Reliance Infocomm launched in 2002 - 2003, it set the calling rate at 40 paise a minute. Going mass market, in the Reliance spirit, he announced is what he wanted to do with broadband as well.

To do this the Reliance team designed a network where consumers used only 10-15% of the bandwidth for voice, allowing Jio, when it was launched, to utilize more of its network for data. Since Reliance was in virgin territory when it came to rolling out 4G in Indian AND was also targeting a mass audience, it also had to address the issue of affordable 4G devices. The company assembled designers from a bunch of different countries including China, Germany, England, Israel and the US to tackle the problem. Adding 4G services to existing phones would make them even more expensive, so the company focused on designing a phone from scratch. The result was the LYF phone, which could be priced as low as Rs. 4,000 (~$55).

From a strategic point of view, this is impressive, but one has to call out RIL's execution and ability to deliver on its vision. Here it is really important to circle back to the core petroleum business. Not many companies in India would have the ability to plough in ~$30B+, the biggest private sector investment in the country's history, to build a broadband network to cover the country. The legacy business gave RIL the buffer and cash reserves to do this (see below for RIL's capex over the past five years).

Source: Morgan Stanley, "Reliance Industries: Annual Report Takeaways"

It also gave RIL the operational experience to do this. Mukesh Ambani's petrochem team specifically had the experience of building, from the ground up, the largest refinery in the world at Jamnagar. Here too Reliance had worked with top notch talent, contracting leading EPC firms like Bechtel, UOP and Foster Wheeler to design and implement the plan. It was pointed out to me that the institutional knowledge of things like acquiring land, licenses, dealing with state and municipal authorities - things that are actually really hard in India - would have been incredibly important at Jamnagar, and would also have served as an important knowledge base as Reliance was building out the infrastructure across the country for Jio.

Digital India

After nearly half a decade of building out the infrastructure to launch its services, Reliance finally formally announced the Jio's launch in 2016. Since RIL had had won the BWA spectrum in 2010 a lot had changed. In 2014, Narendra Modi was elected Prime Minister, with the largest mandate the country had seen since 1984. Since coming to power, Modi had announced a host of government campaigns, including in 2015, "Digital India", to improve online infrastructure and internet connectivity across the country. It was a campaign Ambani tightly tied Jio to. I'll deliver this country's digital future to you, he seemed to be saying to Modi. It couldn't have been clearer when Jio launched. I was in India at the time and remember waking up to each and every newspaper's front page plastered with this ad, which talked about the vision of a Digital India.

Jio was going to make this possible by offering customers free voice calls and zero roaming charges. The data plans were so incredibly cheap, they undercut the incumbents by more than a third. At launch Jio's 4G network already covered 18K cities and towns, 200K+ villages. Within 6 months, the company was promising to cover 90% of India’s population. Ambani said he wanted to get to 100M customers as soon as possible.

Source: Tech in Asia

I still remember the genuine incredulity with which everyone greeted the launch. How was it so cheap? Surely, the network wasn't any good. Surely, your calls would keep dropping. But Jio did work, and, boy, did people flock to it.

Jio's impact on the sector since it first arrived on the scene has been massive. 4G is now the default network for most of the country, and Jio carried ~70% of the country's 4G traffic in 2019. Its competitive pricing in a bid to gain market led to broad reductions in ARPU across the sector, resulting (unsurprisingly) in increased consolidation. Reliance's own share gain has been pretty amazing. It now boasts 388+M customers.

Source: TRAI. Note: Vodafone and Idea merged in Aug 2018. Vodafone share for 2016 includes both entities

Jio continuously innovated even after it launched. For example, to cater even cheaper devices for the mass market segment, Reliance worked to develop the Jio feature phone at the price point of Rs 1,500 (~$20). By last year the JioPhone had ~30% of the feature phone market share. By some estimates the company has sold nearly 100M JioPhones, i.e., ~25% of the subscriber base use the phone. It also adopted a holistic approach to user experience, integrating content with its phones from the get go. It has continued to build out its suite of apps, the most recent being JioMeet.

Source: Morgan Stanley

Jio's growth has been good for Reliance, of course, but also good for the country. Whatever anyone's feelings about the company there's no denying that Jio has really connected India to the internet. Indians consumed ~12GB of data per month in 2019, up from ~90BM in 2014, largely on the back on Jio. That's crazy! We can argue about whether this level of digital connectedness is even a good thing, but that statistic starts giving you a sense of why there has been a frenzy to invest in Jio. If you are an investor who wants exposure to India's untapped potential in the hope that it will bloom, what better investment vehicle could there be?

Where from here?

Watching the Reliance AGM one could almost have forgotten that the company's core business is a staid oil and gas one. More than half the AGM was spent discussing the recent investments Jio (which is now structured as a wholly owned subsidiary of RIL) received over the past couple of months, including the announcement of Google's investment in the company, and the future plans of Jio. Jio is currently only ~9% of RIL's revenue; it will be interesting to see how quickly and how much this share increases.

Source: RIL Annual Report 2019 - 2020

When Facebook announced its investment in Jio there was a lot of talk about the partnership potentially leading to the creation of a super app, and a foray into payments. The cross-sector plays Reliance is now attempting with Jio and retail is interesting, and I'll be interested to see which potential investors come on board on the retail side of the house (Ambani indicated during the AGM that RIL will look to bring strategic and financial partners on board over the next few quarters). But quite honestly, the most tangible immediate partnership that stood out to me out of the recent investments is the one with Google (Jio and Microsoft's cloud collaboration, which has been slightly sidelined given recent raises, is very interesting too). Google announced that it will work with Jio to develop a customized Android operating system to propel a shift of users from feature phones to smart phones (the JioPhone runs of the KaiOS). According to Morgan Stanley, India currently has 300M+ 2G devices (they estimate this segment to contribute ~30% of Airtel and Vodafone’s feature wireless revenues), so there is definitely scope for Jio to steal share here to reach the 500M subscriber goal it has. It will also ensure a deepening of India's internet connectivity.

The timing here is also interesting, given recent tensions with China, and India's dependence on cheap Chinese phones. Could this be a Make in India success story? Maybe. The China angle is important to stress here, because Reliance is nothing if not good at capturing the zeitgeist, sensing political currents and navigating the waters accordingly. My sense is that the company is capitalizing, extremely well, on anti-Chinese sentiment both in India and globally. PE firms are by all accounts sitting on record amounts of dry powder and looking for good opportunities to deploy their funds to get the sorts of returns that are harder to come by with the unlimited money the Central Banks are printing nowadays. For both investors and tech firms, the rapidly digitizing Indian market is one that still presents potential opportunity. China is closed to tech companies, and the US has been threatening to delist Chinese companies from American stock exchanges. In this environment, Jio, with it exposure to India, its scale, and its ability to execute is by all accounts an attractive target.

Reliance is also deftly able to wield the nationalist card when it needs to, to play to a domestic (and even international audience). During the AGM, Ambani talked about an atmanirbhar 5G, which is completely indigenously designed and will be ready to trial as soon as spectrum is made available. He specifically called out the potential to export this solution. Exporting is, of course, firmly in the Reliance wheelhouse, and it is hardly a surprise that Ambani is seizing the geopolitical moment to start laying the groundwork for an exported Jio 5G. The company specifically got a shout out from Mike Pompeo recently for being a "clean carrier". It is building friends in all the right places.

There's an element of this street smarts that also ties back to the launch of JioMeet, which came on the back of the Indian government discouraging the use of Zoom and coincided with India's ban on a whole host of Chinese apps due to security concerns. Jio, which was unabashed in its wholesale lifting of the Zoom UI, offers the service free to Jio subscribers. It also bettered the competition by supporting up to 100 participants and offering uninterrupted “up to 24 hours” long calls. I've seen a bit of sniggering about this on and off Twitter, but again I think this speaks to Reliance seeing an opportunity and tailoring something very specifically for an Indian audience. One use case I recently heard anecdotally was of JioMeet being used for day-long satsangs. This could be an apocryphal story, but to be honest, it is such a specifically Indian use case, that I can believe it. Can you imagine the kind of customer connections Jio can build if it becomes the default satsang app during the era of Covid? I jest, but this is something that is clearly top of mind for the company given its focus on using JioMeet for education and telehealth. Both are solid ways to build real customer connections. Education has the added advantage of potentially hooking customers very early, and ratcheting up the lifetime value you can generate. Given all this am I surprised that JioMeet has been downloaded 5M+ times? No.

The company's strategy and execution has obviously been impressive. But, RIL's broad interests and its tightening hold on India's digital ecosystem also ends up raising questions about whether its size and clout is detrimental to the country. Do I find it disappointing that nearly $20B of foreign capital has gone towards wiping off RIL's corporate debt as opposed to funding start-ups in the country? Yes. But at the same time, Reliance's investments have also helped build the infrastructure that the state clearly didn't have the capacity to, and other corporates weren't bold enough to, build. There is a case to be made for the oversized impact large, disciplined companies can have on areas where they focus their attention. And while we talk about Reliance's oversized presence and think its playbook is marvelous in its simplicity, it's anything but easy to execute on this. Continuing to execute on it will be tougher. It's not like Reliance hasn't had setbacks - JioFibre has been slow to take off, JioChat never gained traction, JioSaavn is a mess etc. And there is a broader organizational question of how Reliance will balance its new, younger digital arm and its more traditional oil business. Will the holding companies be spun out? Jio clearly started off trying to establish a more flat, open office culture but it is still a young company and none of the outcomes are predestined.

There's, of course, also the question of regulatory capture and how much of a role that will play in RIL and Jio's continued success. Pretty much every time an investment in Jio was announced over the last couple of months, a pointed note was made of Reliance's closeness with the current government. These allegations have dogged Reliance regardless of administration. One argument is that Ambani basically controls whichever government is in power. (Perhaps. It is worth noting that the only candidate Ambani openly endorsed during the 2019 parliamentary elections was the Congress candidate for South Mumbai, Milind Deora. Deora is the son of former Petroleum Minister Murli Deora, whose ties to Dhirubhai went back to when they were both yarn traders. That Milind lost the seat despite Ambani's endorsement either speaks to India's democratic credentials or the Congress' utter state of disrepair; I can't decide which.)

I think an easier answer is that Reliance knows how to work the system. I recently read this article by Mark Lutter, which argues that one thing that constrains Silicon Valley's ability to build is that it hasn't engaged seriously with politics. "Part of politics will be co-opting old institutions. Get innovation sympathizers in key positions of power." That is an accusation that could never be leveled at Reliance, no matter which government is in power. This quote from a fantastic 2008 profile of Mukesh Ambani in The New York Times, captures it quite well, I think, ““Remember: these guys all grew up in the License Raj,” said a close friend of the tycoon. “They grew up as lotuses from the filth. It makes them tough, it makes them suspicious, it makes them vindictive at times, and it makes them come out in a hurry. When they were growing up, you didn’t get a second chance.”

During the AGM I don't think Mukesh Ambani left a single government initiative unmentioned: Digital India, Startup India, Make in India, Atmanirbhar Bharat. He checked them all off, and even dedicated the company's 5G to Prime Minister Modi. V.S. Naipaul in India: A Million Mutinies Now, said about Indian politics "there were no principles or party line, there were only personalities...and the relationships were constantly shifting.” This is still incredibly true. It's not like Reliance hasn't been left by the Indian state to fend for itself when it was suitable (for example, blocking the company's attempt to take over Larsen & Toubro in the 1990s or in its on-going skirmish with RIL over income owed from oilfields.) As much as I might dislike admitting this, I think the BJP might be the party of power for a while, and I think the fact of the matter is that RIL probably needs the government as much as the government needs RIL. The relationship is mutually beneficial.

After writing all this, I had to take a step back and remind myself that Reliance, forget Jio, is still a young organization. Think of a corporation as a living organism and Reliance's ability to adapt, acclimatize and mutate to survive stands out. In an interview to Karan Thapar in 2003, Mukesh Ambani noted that most companies rarely last more than one generation. After Dhirubhai's death, he said, he viewed his job as institutionalizing the company and securing its future. What the future holds for RIL and Jio is something we'll see unfold, but I don't think we'd be jumping the gun to say he's gone a long way in securing that future.