The Japanese e-commerce business is still weighed down by its venture into mobile and customer growth has once again slowed.

Hiroshi Mikitani, the boss of Japanese e-commerce company Rakuten, dreamed years ago of imitating Reliance Jio, a newish Indian telco, and leading a telecom revolution in his homeland. Jio had quickly gone from embryo to giant in India after building a modern mobile network and slashing data prices. Hiring Tareq Amin, formerly in charge of Jio’s rollout, Mikitani kicked off an audacious project to build the most automated, hi-tech mobile network anywhere and sell its network technologies overseas. Today, he is taking a samurai blade to costs while his telecom ventures lose money.

Losses, on the plus side, have narrowed. For the recently ended first quarter, the entire mobile segment shrank its operating loss to 71.9 billion Japanese yen (US$460 million), a year-over-year decrease of about 30%. But this was not due to any surge in sales, with revenues up less than 4%, to JPY99.8 billion ($640 million). In the last two years, quarterly operating costs have dropped a fifth, to JPY135 billion ($860 million). After plowing about JPY300 billion ($1.9 billion) into capital expenditure in 2022, Rakuten plans to invest a third as much this year.

The message from the top is that network rollout is largely complete, easing the cost burden on Rakuten. But it has not so far delivered the customer base Rakuten would have wanted or that it now says it needs to break even. This week, Rakuten Mobile, as the domestic network is branded, laid claim to just 6.8 million customers in a country of about 125 million people. NTT Docomo, the market leader, serves more than 80 million.

Subscriber numbers enjoyed a boost in late 2023 when Rakuten added about 810,000 customers in the final quarter after gaining only 330,000 in the third. But growth slowed again to 420,000 in the first quarter of 2024. The impression is of a telco that must constantly dangle generous new offers in front of the Japanese to maintain their interest, which sometimes proves short-lived (although churn is also down). Rakuten estimates it needs between 8 million and 10 million customers to stem losses.

Basestation spotting

The danger for investors is that relentless cost cuts hinder Rakuten’s ability to go much beyond these targets. For all the talk of network rollout being largely finished, Rakuten has yet to bring new 700MHz spectrum into use. That’s needed to plug important coverage holes and will necessitate the installation of new 700MHz radio heads on masts, next to the 1.7GHz radio heads already in service.

Work is to start this summer, with Rakuten estimating the cost over a long ten-year period at roughly JPY56 billion ($360 million). How can it be so low? “Because we are manufacturing this at a new type of factory – that’s the answer,” said Mikitani, without providing further details. Nokia, though, is the vendor that landed the contract to provide these radios. Judging by Mikitani’s description, it is being forced to produce these on a very tight budget.

The 700MHz rollout is not the only network expansion job ahead of Rakuten. Basestation spotters could previously delve into the annexes of earnings presentations and read details of site numbers for both 4G and 5G. Those figures revealed a huge gulf between the two generations when they were last shared at the end of 2023, with about 61,000 4G basestations in service against as few as 11,600 5G ones. Rakuten stopped disclosing the numbers this quarter while promising to expand 5G coverage in important areas.

The reasonable excuse for this 5G lag is that satellite services are still interfering with the relevant frequency bands. “So even if we deploy sites, we cannot transmit at full power,” said Sharad Sriwastawa, Amin’s successor as CEO of Rakuten’s telecom ventures, when Light Reading caught up with him at Mobile World Congress earlier this year. He also conceded that a 5G rollout would have implications for the annual capex forecasts shared by Rakuten.

These, however, have not been raised from the JPY100 billion ($640 million) level for this year and next. Despite that, Rakuten is now targeting a 60% expansion of 5G coverage across swathes of Greater Tokyo by the end of this year, compared with the level seen in January. And unlike other Japanese telcos, Rakuten has based its 5G rollout so far on massive MIMO, a high-performing but costlier technology. This may deliver service advantages when built. But the building of it could be expensive.

Symphony’s bum notes

Rakuten Mobile can at least see a path to profitability, assuming customer growth is not upset by efficiency measures. The bigger worry is Rakuten Symphony, the unit that competes against other vendors of network products on the world stage. In 2021, Rakuten estimated Symphony’s addressable market was generating about $90 billion (at the midpoint of its range) in annual revenues. By 2025, it would be worth $140 billion, said Rakuten.

But these figures and forecasts today appear ludicrous. Much of the projected growth, for instance, would supposedly come from the market for radio access network products, where Rakuten was targeting a 25% share. Yet this market shrank 11% in 2023, according to Light Reading sister company Omdia, which is guiding for another decrease of between 4% and 6% this year. Symphony made $476 million in total revenues in 2022, compared with the $17.8 billion generated by Ericsson’s networks unit alone. Last year, Symphony’s revenues fell to $393 million.

Back in Japan, the cost cutting seems unlikely to end soon. “With AI, we’d like to further reduce numbers,” said Mikitani in reference to monthly operating costs at Rakuten Mobile. If nothing else, it proves networks can be automated to a level unimaginable a few years ago. If it can finish the work on 5G, and persuade a few million more Japanese customers to switch from their old providers, it might just have a profitable future ahead.

Original article can be seen at: