After Free and RJio, Japan’s Rakuten prepares to wreak disruptive havoc

There can be few business prospects more unappealing than entering a saturated and competitive mobile market as a new entrant, without any of the infrastructure, brand awareness and partnerships enjoyed by the incumbents. Yet, even as established operators in many markets look to consolidate, and achieve better economies of scale, there are still some bold companies which believe they can change the rules of the mobile game, and come out on top.

Some of these are established in adjacent markets – wireline providers moving into the mobile space for instance. That is challenging, but they do have the advantages of established customer bases in the telecoms and media space, plus a wealth of home WiFi and wireline backhaul links to improve their economics. Comcast and Charter in the USA, Free Mobile in France, TPG in Singapore – these are companies which leverage WiFi heavily to add wireless to their offerings, and are now deploying their own cellular networks (in the US cablecos’ cases, that will happen soon, in shared spectrum or 5G).

Even greater challenges await companies moving into cellular without a wireline base, as Reliance Jio initially did in India (though it is now acquiring fiber assets); as Free’s parent Iliad is doing in Italy; and as ecommerce player Rakuten is doing in Japan. For these brand new entrants, it will be essential to make a virtue out of the lack of existing physical assets; to achieve a radically lower cost base than the incumbents can hope for, often by relying on partnerships; and to launch differentiated services that will be hard to emulate.

Rakuten, which already operates a low cost MVNO service in Japan, announced in December 2017 that it would enter the market as a full MNO. It claimed this would enable it to deliver higher value, more profitable services and seize more market share, while also having control over the quality and behaviour of its network. The company’s president Mikitani Hiroshi said at the time: “We will develop our MNO business into a driving force like our credit card business.”

In April, the company was allocated 2×20 MHz of 1.7 GHz spectrum plus unpaired airwaves in 3.4 GHz, with the Japanese Ministry of Internal Affairs and Communications saying it wanted to stimulate more competition, and lower prices, in a market dominated by the three big players – NTT Docomo, KDDI and Softbank. Rakuten will not launch commercial services for another year, but aims to reach 96% of the population within six years, and to sign up 10m subscribers by the end of fiscal year 2028. Its plan is to invest about ¥526bn ($4.7bn) in that roll-out, over five years, a figure it has already reduced from an initial estimate of ¥600bn ($5.4bn).

By contrast, NTT Docomo said in its most recent financial results announcement that it would spend ¥200 ($1.8bn) a year for the next five years on 5G alone. The telco’s total capex spend in the 2018 fiscal year (to March 31) was ¥576.4bn ($5.1bn). This shows how a new entrant can (and must) deploy with a far lower cost base.

After all, Docomo is not an extravagant telco itself. It has persistently made it clear that it will only deploy 5G at significantly lower capex than it spent on 4G, and reiterated the message on its results call last week. Rémy Pascal, an analyst at Analysys Mason, said the ¥1 trillion 5G figure “will not be linear but roughly one third of their capex over the period”.  That calculation means Docomo’s total capex would amount to about ¥3 trillion ($26.6bn) in the period, an average of about ¥600bn ($5.3bn) a year – which, means 5G would only add 4% to its spending, compared to its just-completed fiscal year. The Japanese operator’s overall capex fell by 3.5% in its last fiscal year, to ¥576.4bn  ($5.1bn).

At February’s Mobile World Congress, Docomo CTO Onoe said one of the “killer services” for 5G might just be the ability to boost data capacity “with no increasing trend in capex”. His presentation showed that the MNO’s capex has been in steady decline since the launch of 3G in 2001, when it peaked at almost JPY1 trillion ($9.3bn), while in 2017 it was JPY600bn ($5.6bn), and the operator is not projecting significant increases for 5G. “In the past capex has been on a downward trend,” said Onoe. “We can continue this trend.”

Despite this optimism about the affordability of 5G, The big three MNO, as well as the main vendors, have said Rakuten’s estimated capex will be too small to build a wireless network to cover even Greater Tokyo, let alone the other main metro areas of Nagoya and Osaka (these three heavily populated regions are the new operator’s first targets).

But a senior official of the Ministry of Internal Affairs and Communications, in a recent interview with Japanese newspapers, said a fourth MNO is necessary to launch the “cooperative oligopoly “ of the other three, and hinted that there might even be grants or subsidies to help that happen. In particular, he said that, if Rakuten can only raise sufficient funds, in the first phase, to offer limited coverage, the regulator could compel the incumbents to allow the newcomer to use their networks for nationwide roaming at an affordable rate. Then Rakuten can expand its coverage at the appropriate rate for its business, even if that takes decades to complete, he said.

However, Rakuten is not relying on state support, but on creative partnerships to make its capex and opex dollar go further.

Just as Reliance Jio accelerated its 4G-only roll-out, and reduced its capex and opex costs, by signing partnerships with site, fiber and spectrum owners, so Rakuten has formed highly strategic alliances with utilities in Japan. It will use their power transmission towers  and power distribution pillars, and other sites, plus some of their fiber, to support its deployment, along with conventional cell towers and rooftops. Its partners include Hokuriku Electric Power Company, Kansai Electric Power, Chubu Electric and TEPCO Group, and it may seek other local utility deals. This is not a conventional asset leasing agreement – instead, the power companies get priority, and subsidized, access to Rakuten’s future 4G and 5G networks to support their smart grid and other connected activities.

This is a co-investment pattern which is starting to interest many operators round the world as they look to address two challenges which will often be in conflict – reducing the total cost of ownership of the network in the 5G era; and optimizing that network, not just for mobile broadband, but for the demands of various vertical industries, each with its own requirements in terms of bandwidth, latency, availability and security. Some are exploring models whereby those verticals share the cost burden, in return for getting the kind of network they want at an affordable rate. The Japanese MNOs have worked with railways and cities in this way, while China Unicom underwent a significant restructuring, which involved Internet and industrial companies taking stakes in the ailing MNO, in return for influence over the shape of its 5G network, and priority access to it.

In another strategic pact, it is actually working with one of its future competitors, KDDI. Rakuten will help the second Japanese MNO launch a mobile payments service, in return  for the use of KDDI’s 4G network to extend coverage via roaming. NTT Docomo has been a trailblazer in mobile payments for a decade but Rakuten’s successful experience with its own financial services unit could help KDDI narrow the gap.

KDDI will “provide roaming services to Rakuten for its 4G mobile network”, the companies said in a joint statement, enabling the newcomer “to offer a nationwide LTE service from launch”. The services will be provided until March 2026 which will give the ecommerce giant time to build out its own network. Meanwhile KDDI will use Rakuten’s payment platform and network of around 1.2m affiliated stores in Japan to launch its own barcode and QR payment service, called au PAY, in April 2019.

“By promoting the mutual use of both companies’ payment services, the companies aim to improve customer convenience and accelerate their vision of a cashless society,” Rakuten added. Rakuten will also provide its logistics services to KDDI’s Wowma online shopping channel from April 2019.

Of course, KDDI shareholders will pray the operator does not come to regret its decision, in the same way that Orange did in France when it hosted Iliad’s Free Mobile launch, allowing Free to start with a limited build-out of its own and roam on the incumbent’s network.

French regulator Arcep came under fire from the other MNOs, SFR and Bouygues, for pushing Orange to support roaming for the disruptive new competitor, and in 2016 the incumbent ended that deal, though it will not be entirely phased out until 2020. The damage was done for the older operators – the roaming arrangement allowed Free to start introducing cut-price offers, even though it initially only had about 3,000 base stations built out. Today the end of the deal in 2016, it had almost 12,000, still a small number compared to Orange’s 50,000 multi-radio base stations. But 12,000 base stations are capable of catching a very large percentage of calls (around 78%).

Free had demonstrated to the world that, by using the WiFi in its 5m home gateways installed throughout France, it could offload at least 70% of internet data to its own WiFi network, greatly reducing its data usage on the Orange network (saving its roaming fees) or its own need to boost cellular capacity. Even so, it was said that it paid Orange over €1bn a year in roaming charges.

Helped by the ability to build out gradually, Free grew, from its launch in 2010 until 2016, to 11.9m mobile customers or around 17% market share, having unleased a savage price war  – its cheapest phone deal is €2 a month- and sparking a wave of consolidation, including the acquisition of SFR by Iliad’s rival Numericable.

Currently, NTT Docomo hosts Rakuten’s MVNO, though the deal with KDDI suggests that the second Japanese operator will be the main partner for its 4G MNO deployment. That prospect will only redouble the incumbents’ efforts to reduce costs and avoid the fate of their French counterparts.

One area where many incumbents have an advantage in this respect is in existing physical infrastructure. Docomo can be fairly thrifty in the 5G capex period is that it has already invested significantly in infrastructure including fiber for multihaul and new sites. Its parent group owns significant physical assets and where it needs additional capacity for virtualized RAN fronthaul, dark fiber is cheap in Japan.

THis shows how all operators need to leverage existing physical assets – fiber, sites, in-home WiFi routers and so on – to make 5G roll-out more cost-effective and faster. For new entrants, this is critical to their ability to present a challenge to the incumbents at all. Rakuten and Reliance Jio are looking better positioned than most new entrants because they have been able to form large-scale strategic alliances with infrastructure owners, of a kind that are tough to achieve for most new players.

Even so, the model of leveraging a low cost base – one that existing operators will not be able to emulate because of all their legacy equipment and customers – will only go so far for new entrants. The Reliance Jio case study shows how a new challenger may use extensive physical assets and a cost-effective model to gain scale quickly with cut-price offers – but that needs to be followed up with higher value, differentiated services, or the new player is likely to go the same way as many low cost MVNOs, either failing, or being snapped up for a low price by an existing MNO.

For Rakuten, it will be important to make the transition to higher value quickly.

For one thing, operators facing similar new entrants have often reduced their fees pre-emptively. In India, even the threat of Reliance Jio’s launch sparked off a price war which continued after it entered the market with an introductory offer of free data and voice. In Italy, all three MNOs introduced low cost tariffs in anticipation of Iliad’s launch there. The same seems to be happening in Japan. Reports last week indicated that Docomo was preparing to reduce mobile prices some months before Rakuten’s expected launch date, and the other two would likely follow suit.

For another, there will be pressure on the fledgling MNO to deliver better operating margins than the average mobile provider, since the group’s profits are under pressure. In its core business, ecommerce, it is challenged by Amazon and the unit’s operating profit fell by 3.8% in the most recent fiscal year. The organization has been investing in other businesses to offset that decline, including financial services (credit cards and Internet banking), whose operating profit rose by 11% in the fiscal year, helping drive the group to its first net profit increase for three years.

Similar targets will be set for the mobile business after its full commercial launch next year and, as with the finance operation, Rakuten will also look to expand its overall customer base by offering new services, and then add to its rich store of customer data, one of its most valuable assets. The Rakuten Mobile MVNO already runs a loyalty points scheme which it said contributes valuable insights into consumer behaviour, which can be sold to marketeers and advertisers.

Another physical asset which will help its launch marketing is the national network of 218 stores which support its existing Rakuten Mobile MVNO services – that not only gives it a retail presence, but an established brand, although one challenge will be to change consumers’ perception of that low cost brand as it moves into higher value services.

RJio showed how to launch an attractive offer with a very efficient cost base. However, it has mainly relied on free or low cost voice and data to achieve its rapid customer build-up, and this is in a market where ARPUs and margins were already tiny. So in the past year or two, it has moved beyond being just a cost cutter, and has been building a full portfolio of digital services, investing in content as well as smart home, connected car and mobile money offerings. That will force a response from its rivals, which could provide interesting new choices for consumers, but in the short term, pile even more capex burdens on the hard-pressed MNOs.

RJio is working with Eros International Media to produce and consolidate content from across the country and says it will invest up to INR10bn ($150m) to produce and acquire Indian films and original digital content across all the country’s major languages. It has also signed a deal to integrate Saavn, an over-the-top music platform, with its own JioMusic service.

RJio also has a 50% stake in Viacom 18, the parent company of Colors Channel, a popular TV channel, as well as a 25% stake in Balaji Films. And it is looking to acquire one or two wireline operators to increase its already substantial base of fiber to support its fixed broadband services and its network densification.

Despite its investments, its low overall cost base and rapid growth in market share enabled it to become profitable more quickly than many expected. For its financial year, ended March 31, it reported earnings of INR7.23bn ($108m) and a second consecutive quarter in profit.

But it will continue to have to strike a very difficult balance between improving its customer proposition and keeping its network costs under control. This will be hard, given that its average monthly data consumption per user hit 9.7GB in its fourth quarter, the highest in India and one of the highest rates in the world, driven by its low tariffs, and the high number of Indians who use their handset as their primary Internet connection – through choice or because they have limited access to fixed broadband or even TV. Average video consumption per user per month was 13.8 hours.

Japan is a very different market to India, but both have very high rates of mobile usage. Rakuten will be facing many of the same dilemmas as RJio as it moves ahead with its 4G and 5G launches over the coming year.

Rakuten is already planning for 5G:

Rakuten is initially deploying 4G, taking advantage of the falling costs of this mature technology. Like Reliance Jio, it will also leverage its greenfield status by adopting low-opex, software-driven architectures from day one. RJio devised its own end-to-end SON (self-optimizing network) technology as a step towards network automation that goes a lot further than its rivals. Rakuten has spoken of similar approaches, including a high level of virtualization, and it also plans to tap into the experiences of the very price-sensitive Indian market, especially as it starts to plan for 5G.

It will be under pressure to have a short term 5G roadmap, since the Japanese MNOs are set to start rolling out the next generation services from late 2019. However, it will be tougher to get the same cost efficiencies, in the early days, from a new technology which will still have some degree of price premium, and will often require significant integration expertise in order to deploy new architectures, including open cloud platforms, effectively.

Rakuten Mobile has signed a memorandum of understanding with Indian integrator Tech Mahindra, which is very active in some of the 5G-focused open network initiatives like Facebook’s Telecom Infra Project and the Open Networking Foundation’s CORD. The partners will cooperate in building 4G and 5G labs in  Tokyo and Bengaluru, India, and pledged to create “a 5G-ready network lab which will be one of its kind in the industry”.

In June, Rakuten announced its first 5G trials, conducting over-the-air tests in the 28 GHz band with Nokia and Intel.

MNOs insist 5G will not cause a capex bubble:

At the same event, US Cellular’s CTO, Michael Irizarry, said 5G “would have to cost no more than 4G and probably less.” That isn’t too far from what Vodafone’s CTO, Johan Wibergh, has said in various forums – that 5G has to justify itself on the basis of cost efficiency alone, before new revenue streams are considered. At MWC, he reiterated his theme, saying: “Revenue is not growing and has been very flat or rising just 1%, and so there is no room for increasing costs.”

At the 5G World Summit in London earlier this year, Swisscom and Three UK also played down hopes of a 5G-related boost in capex, largely because they do not believe 5G will create a significant boost in spending by their own mobile consumers. Three UK’s CTO, Bryn Jones, insisted – in a clear message to suppliers – that he saw no reason for a major capex increase, saying: “Networks are like painting the Forth Bridge. You roll out 2G and as soon as that is rolled out you start rolling out 3G. The capex is really substitutional and you manage within the envelopes.”

His counterpart at Swisscom, Heinz Herren, also said 5G could be managed within today’s “capex envelope”, adding: “With 5G there might be some incremental capex but if we are clever and add radios that can deal with 4G and 5G we will not see a big uplift in capex.”

He added that the business case still had to be made for at-scale roll-out, saying: “The Swiss market is flat and so it is really difficult to add 5G and sell it and personally I think we are having too many discussions around the business case. I think if your main business case is connectivity there is no question about doing 5G, but I don’t think you will see additional ARPU.”

Among the reasons why operators believe they can deploy 5G without raising capex are price wars in base stations; the virtualized, software-driven architectures in which it has been an early mover; and more open platforms.

NTT Docomo recently boasted of its early stage gains from using commodity servers as it moves towards fully virtualized networks. Although this transition is far from complete, Docomo says it has already achieved 10% reduction in relevant capex.

Hiroyuki Oto, general manager of the MNO’s core network development department, said the commodity hardware had resulted in capex savings despite some additional costs associated with management and network orchestration.

Constantine Polychronopoulos, CTO at virtualization company VMware, said: “We are moving toward software-defined radios and eventually upgrading a radio will just be a software exercise and a lot less costly. Virtualizing the network will address capex as well.”