Rakuten Group: long-term performance & turning pointsSales (revenue) and profit-margin ratio
Sales (¥ bn)Net margin (%)
1997From a ¥50,000-a-month virtual mall to a three-domain business base
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
FY2001 · consolidated
Revenue$55M
Net income—
Net margin—
→
FY2004 · consolidated
Revenue$421M
Net income—
Net margin—
1997M.D.M. Inc. incorporated in Minato-ku, Tokyo, with ¥10m capital
1997Rakuten Ichiba opens in May with 13 shops at ¥50,000 a month
1999Renamed Rakuten, Inc.
2000Registered over the counter with the Japan Securities Dealers Association
2000Infoseek acquired outright for ¥9bn after Yahoo! Shopping takes first place
2001Rakuten Travel launched
2002Mall fees move to usage-based charging in April
2002Rakuten Super Points launched
2003My Trip Net (Tabi no Mado) becomes a subsidiary
2003DLJdirect SFG Securities becomes a subsidiary
2004Aozora Card acquired for ¥7.4bn
2004Tohoku Rakuten Golden Eagles approved as a new club; JASDAQ listing in December
Rakuten's first seven years were an argument about price. Mikitani Hiroshi opened a mall that asked shopkeepers for a twentieth of what the department-store malls charged, rode the resulting rush of merchants to an over-the-counter listing, and then — with the business at its best — tore up the very flat fee that had made him, rebuilding the company around a share of what its shops sold. The cash that came out of that change, and out of the offering, went straight into buying travel, securities and card businesses, so that the mall of 1997 ended 2004 as three businesses under one membership ID and a professional baseball club.
A flat ¥50,000 a month, and the 6,150 shops it drew
In February 1997 Mikitani Hiroshi (三木谷浩史), thirty-one years old and newly resigned from the Industrial Bank of Japan, incorporated M.D.M. Inc. with capital of $82,631 (¥10m) at Atago in Minato-ku, Tokyo, and in May of the same year opened Rakuten Ichiba. Thirteen shops were trading on the day it opened. Where the virtual malls run by department stores took listing fees in the order of $8,263 (¥1m) a month, Rakuten set a flat $413 (¥50,000) a month — an order of magnitude lower. An editing tool called RMS, which required no programming, let shopkeepers build their own storefronts; the target customers were the individual proprietors of local shopping streets, and the threshold for opening a shop was lowered to suit them. The contract required six months of system fees to be paid in advance in a single sum, which funded working capital without recourse to outside money and secured recurring profits in two consecutive years from the financial year ended December 1998. The first year's shops were won the hard way, by sales staff calling on shopping streets one door at a time.
This pricing brought a surge in the number of shops: 320 by December 1998, 1,800 by December 1999, 4,800 by December 2000 and 6,150 by January 2003. Listing-fee revenue accumulated in direct proportion to that count, and until this point the earnings of the mall-only business rested on it continuing to rise. In June 1999 the company changed its name from M.D.M. Inc. to Rakuten, Inc. In April 2000 it registered over the counter with the Japan Securities Dealers Association and raised $459.4M (¥50bn) in the offering, against sales of only $10.5M (¥1bn) for the January-to-June period of that year. Even through a continuing consumer slump, gross merchandise sales on Rakuten Ichiba grew from $430.4M (¥52bn) in 2001 to $598.7M (¥75bn) in 2002, a rise of about 40 per cent, reaching a scale comparable to a department store in Ginza, Tokyo.
The ¥9bn Infoseek acquisition, and Yahoo! closing in
While the use of the $459.4M (¥50bn) raised over the counter was still undecided, managing director Yamada Yoshihisa (山田善久) said, buying is easy, but money is not something you should spend just because you have it (Nikkei Business, 23 Oct 2000), and was in no hurry to deploy it. Then, in October 2000, a survey by Nielsen//NetRatings of the United States showed Yahoo! Shopping overtaking Rakuten Ichiba in user numbers to take first place. With its first-mover position under threat, Rakuten announced at the end of November 2000 that it would acquire the loss-making portal operator Infoseek outright for $83.5M (¥9bn). For a company that had made a point of profits, taking on a business carrying a recurring loss of $6M (¥650m) over the year from October 1999 to September 2000 was a change of line. Mikitani described the aim of the acquisition as to become an integrated media company with shopping at its core (Nikkei Business, 11 Dec 2000), signalling that the money raised in the offering would go on closing the gap with Yahoo! in the power to draw traffic.
At Infoseek after the acquisition, Mori Manabu (森学) — not a man from that company but an ordinary Rakuten employee — was picked out to lead the turnaround, and took a business with an operating loss of about $12.1M (¥1bn) in the year to September 2000 to an operating profit of $375,160 (¥47m) in the June-to-September quarter of 2002. Even so, the dependence on the number of shops — lose the power to draw traffic and the merchants who were the source of revenue would leave — had already been pointed out at the end of 2000. In March 2001 the company opened Rakuten Travel, beginning to diversify away from a single shopping business. From 2001 the growth in shop numbers slowed, and commentary questioning the revenue model itself — one that depended on the shop count — became conspicuous.
From “property” to “franchise”, and the three-domain structure completed
In February 2002 Rakuten announced that it would revise the flat $399 (¥50,000) fee it had charged since 1997, moving to a usage-based charge that added a system fee for shops selling more than $7,982 (¥1m) a month. It was a switch from a property model, letting space at a fixed rent, to a franchise model that earned in step with the merchant's sales, and Mikitani himself took the front line, doubling as head of the sales division from July 2002. The change had been preceded by roughly a year of simulation by a special team, and Mikitani described the judgement of the time: from the founding I thought the flat-fee system would have to be changed sooner or later. I was prepared to see it through even if the number of shops halved (Nikkei Business, 10 Feb 2003). The sales organisation was rebuilt from winning new merchants to consulting for existing ones; one e-commerce consultant covered around 150 shops, and about 60 of the company's roughly 290 employees were put on the work.
The new fees applied from April 2002 and were first charged in May; in that month departures exceeded openings by 58 shops, but from June the net figure rose again. Consolidated sales for the year ended December 2003 were $156.1M (¥18bn) and recurring profit $38M (¥4bn), close to double the previous year. In this period Mikitani had set a medium- to long-term target of $862.7M (¥100bn) in recurring profit — the level of Japan's top thirty companies outside the financial industry. In September 2003 the company took control of My Trip Net, operator of the accommodation-booking site Tabi no Mado (旅の窓口), and in November of DLJdirect SFG Securities (now Rakuten Securities), spending a little over $517.6M (¥60bn) in a short span to bind shopping, travel and finance together. Total transaction value reached $364M (¥42bn) in the October-to-December quarter of 2003, up 76 per cent on the same quarter a year earlier.
In September 2004 Rakuten paid $68.4M (¥7bn) for all the shares in Aozora Card, the card-loan company held by Aozora Bank and Orix. Rather than lean on the co-branded issuance with Sumitomo Mitsui Card begun that August, it meant to run a card business of its own. The company it bought was a small loss-maker — incorporated in December 2001, with a loan book of $182.1M (¥20bn) and 23 employees at the end of March 2004 — and the price was modest beside the $278.6M (¥32bn) for Tabi no Mado or the $285.5M (¥33bn) for DLJ Securities. In that same year of 2004, with professional baseball reorganising, Rakuten entered the game. On 24 September it applied to join the Nippon Professional Baseball Organization, setting out a plan for a new club based in Sendai, Miyagi prefecture. The review came down to a head-to-head with Livedoor, and the deciding factor was the difference in earning power: against Livedoor's recurring profit of $47.2M (¥5bn), Rakuten expected $134.1M (¥15bn) for the year ending December 2004, and weight was given to its capacity to absorb losses that averaged $29.6M (¥3bn) across the five Pacific League clubs. The owners' meeting of 2 November approved the entry of the Tohoku Rakuten Golden Eagles unanimously, and the first new club in fifty years, since 1954, was born. That December Rakuten listed on the JASDAQ Securities Exchange, moving from over-the-counter registration to a formal market.
2005Pulling finance inside — the ecosystem widens, and the first loss since founding
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
FY2005 · consolidated
Revenue$1.2B
Net income$176M
Net margin15%
→
FY2018 · consolidated
Revenue$10.0B
Net income$1.3B
Net margin12.9%
2005Total assets jump from ¥307.5bn to ¥1.66tn as finance is consolidated
2007Fusion Communications becomes a subsidiary
2008First net loss since founding, on Lycos Japan and overseas impairments
2008¥20bn of preferred shares underwritten in the troubled eBank
2009eBank consolidated as a subsidiary; renamed Rakuten Bank in 2010
2010bitWallet acquired, bringing Rakuten Edy into the group
2010English declared the official internal language
2011Net loss on an ¥84bn impairment in overseas e-commerce
2012Kobo acquired; Airio Life Insurance made a subsidiary
2014Viber Media and Ebates acquired; Rakuten Mobile starts as an MVNO
2013The Golden Eagles win the Japan Series; listing moves to the TSE First Section
2018Asahi Fire and Marine Insurance made a subsidiary
The years after 2004 turned Rakuten from a mall operator into a balance sheet. Securities, cards and then a bank came inside the group, multiplying total assets more than fivefold in a single year and giving the company a second pillar of earnings that eventually rivalled merchandise. The same appetite ran overseas, where roughly $1.9B (¥200bn) of acquisitions bought reach in Europe and North America and, twice, wrote it off again — the first loss since the founding in 2008, another in 2011 — while at home a common pool of points quietly turned a collection of purchases into one economy.
From ¥307.5bn to ¥1.66tn — a balance sheet remade by finance
With the three-domain structure established, consolidated total assets swelled from $2.8B (¥308bn) in the year ended December 2004 to $15.0B (¥1.66tn) in the year ended December 2005. It was the consequence of taking credit, payments, securities and banking inside the group: by the year ended December 2006 the finance segment alone held $3.6B (¥417bn) of assets in credit and $4.5B (¥528bn) in securities. Consolidated sales also grew sixfold, from $1.2B (¥130bn) in 2005 to $7.2B (¥782bn) in 2016. Against that, impairment on the acquired Lycos Japan and on overseas television-shopping operations produced an extraordinary loss of $782.9M (¥81bn) in the year ended December 2008 and a net loss of $531.3M (¥55bn) — the first loss since the founding. Sales that year were $2.4B (¥250bn) and operating profit $456.8M (¥47bn), so the core business stayed in the black; the loss was the clearing-up of investments.
In 2008, the year of that first loss, Rakuten underwrote a capital increase at the troubled eBank. The bank had posted a net loss of $198.7M (¥23bn) in the 2007 financial year, and its capital adequacy ratio had deteriorated to 6.4 per cent at the end of June 2008; of a capital raising of up to $387.1M (¥40bn), Rakuten carried $193.6M (¥20bn) in preferred shares. For a company that had been running securities and credit, this was its first investment in a bank. In February 2009, having obtained the Financial Services Agency's approval as a major shareholder, it converted the preferred shares into ordinary shares and consolidated the bank as a subsidiary with 48.69 per cent of the voting rights. The bank changed its name to Rakuten Bank in May 2010 and became wholly owned through a share exchange that October.
In January 2010 Rakuten acquired bitWallet, bringing the electronic money Rakuten Edy into the group. In telecommunications it made the IP-telephony operator Fusion Communications (now Rakuten Communications) a subsidiary in August 2007, and in insurance it took control of Airio Life Insurance (アイリオ生命保険, now Rakuten Life Insurance), until then an equity-method affiliate, in October 2012. From the year ended December 2013 it moved to International Financial Reporting Standards and reported in two segments, internet services and internet finance. Operating profit in the FinTech business reached $601.8M (¥66bn) in the year ended December 2016, and finance became a pillar of earnings standing alongside merchandise.
The goodwill from ¥200bn of overseas acquisitions, and the 2011 net loss
Through the first half of the 2010s Rakuten made overseas acquisitions one after another: PriceMinister of France in July 2010, the e-book company Kobo in January 2012, the mobile-messaging company Viber Media in March 2014 (on the order of $850.3M (¥90bn)), and the North American online cash-back service Ebates in October of the same year (on the order of $944.8M (¥100bn)). Alongside these it acquired Wuaki.TV, a Spanish video-streaming operator (now Rakuten TV Europe), in June 2012 and Viki in September 2013, gathering up digital content in Europe and North America. In 2010 Mikitani declared English the company's official internal language, holding out a conversion into a global firm. In the year ended December 2011 impairment on the overseas e-commerce business produced an extraordinary loss of $1.1B (¥84bn) and tipped the company into a net loss of $13.8M (¥1bn).
PriceMinister's brand was closed in 2021, while Ebates continued as Rakuten Rewards in North America. Interest-bearing debt swelled more than threefold, from $4.0B (¥390bn) in the year ended December 2013 to $11.2B (¥1.23tn) in the year ended December 2018. At home, the Tohoku Rakuten Golden Eagles won the Japan Series for the first time in November 2013, and 214,000 people gathered for the victory parade through Sendai. That December the company moved its listing to the First Section of the Tokyo Stock Exchange. In June 2017 it established Rakuten LIFULL STAY (now Rakuten STAY) to enter the private-lodging business.
The “no-touch model”, and the ecosystem idea taking hold
From the founding years Mikitani had made the inventory-free business model the axis of what he chose. Neither travel nor securities carries physical distribution with it; Rakuten Ichiba held no stock; and the book inventory behind Rakuten Books was left entirely to the wholesaler Nippan. Mikitani said plainly, I like the “no-touch model”, where you hold no inventory (Nikkei Business, 22 Mar 2004), and went further: I think a company's value is ultimately decided by ROE. In the end only two things matter — whether the company has social justice, and whether it is distributing returns to shareholders (Nikkei Business, 22 Mar 2004). He set a target of $9.2B (¥1tn) in combined gross transaction value across Tabi no Mado and the shopping mall, and said that with the moves into finance and travel a large part of the acquisition puzzle had been filled in.
Rakuten Super Points, begun in November 2002, became a common currency of cross-use — points earned on Rakuten Ichiba could be spent on travel, securities, cards and banking — and, with the membership ID binding the businesses together, formed the prototype of the Rakuten ecosystem. In 2005, the year it took on a baseball club, advertising expenditure rose to $71.7M (¥8bn), close to four times the previous year, and most of it is thought to have been related to the Tohoku Rakuten Golden Eagles. Against the $282.8M (¥33bn) of advertising spent over the five financial years to December 2007, market capitalisation grew from $726.4M (¥91bn) at the end of 2002 to $6.1B (¥719bn) at the end of 2007, and the company was rated first for the growth in market value it obtained per yen of advertising. The year ended December 2017 set new records, with revenue of $8.4B (¥944bn) and operating profit of $1.3B (¥149bn). In March 2018 it made Asahi Fire and Marine Insurance (now Rakuten General Insurance) a subsidiary, completing a line-up of cards, banking, securities, life insurance and general insurance.
2019The fourth mobile carrier, and more than ¥840bn of net losses
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
FY2019 · consolidated
Revenue$11.6B
Net income-$292M
Net margin-2.5%
→
FY2025 · consolidated
Revenue$16.7B
Net income-$1.2B
Net margin-7.1%
2017Board resolves to enter mobile as the fourth MNO in December
2019Rakuten Mobile begins carrier service in October
20205G service launched in September
2021Renamed Rakuten Group, Inc.; Rakuten Symphony launched in August
2022Population coverage passes 96 per cent; move to the TSE Prime Market
2022Record net loss of ¥372.8bn; alliance with Mizuho Securities in October
2023Rakuten Bank listed separately on the TSE Prime Market
2023The Rakuten Saikyo Plan launched; 700MHz platinum band awarded in October
2023Rakuten Payment and Rakuten Card merged in November
2024Rakuten Card and Mizuho Financial Group enter a capital and business alliance
2025Rakuten Mobile reaches EBITDA profitability in the first quarter
2025Total contracted lines pass ten million in December
In 2019 the company that had prided itself on owning nothing began pouring concrete. Rakuten switched from renting other operators' lines to building its own base stations, betting that a fully virtualised network would make the arithmetic work, and instead ran up losses on a scale that shook the whole group's finances. What saved the bet was the willingness to bend it — roaming, platinum band spectrum, an alliance with Mizuho, and finance profits large enough to carry the telecommunications loss until the first quarterly profit arrived in 2025.
Why build its own base stations — the December 2017 announcement
On 14 December 2017 Rakuten's board resolved to enter the mobile telephone business as a carrier (MNO). The plan was to break into a market held between NTT Docomo, KDDI and SoftBank as a fourth MNO, with service starting during 2019, more than 15 million subscribers, and capital expenditure of about $1.8B (¥200bn) at the outset and up to $5.3B (¥600bn) by 2025. It had been running as an MVNO on the three incumbents' lines under the Rakuten Mobile name since October 2014, and it aimed to become an operating company delivering e-commerce, FinTech and telecommunications as one on that subscriber base. At the time of the announcement Rakuten IDs numbered about 100 million and cumulative issuance of Rakuten Super Points had passed $8.9B (¥1tn); the company set out to use both to integrate the ecosystem.
The other bet was the wholesale adoption of a fully virtualised network (Open RAN), running base-station software on general-purpose servers. Mikitani called this the Apollo programme of the mobile industry and argued that it could cut infrastructure costs by 70 to 80 per cent against the existing operators. To sell the technology on, he launched the telecommunications platform organisation Rakuten Symphony in August 2021. Against that, because it opened for business without an allocation of platinum band spectrum (the 700MHz band), the network carried a weakness in signal quality indoors and underground.
¥840bn of accumulated losses; total assets from ¥7tn to ¥20tn
Rakuten Mobile began carrier service in October 2019, and base-station construction started in earnest. Net losses widened — $292.6M (¥32bn) in the year ended December 2019, $1.1B (¥114bn) in the year ended December 2020 and $1.2B (¥134bn) in the year ended December 2021 — and the year ended December 2022 recorded the largest ever, a net loss of $2.8B (¥373bn). Net losses from 2019 to 2023 exceeded $6.0B (¥840bn) in total. Interest-bearing debt swelled from $11.2B (¥1.23tn) in the year ended December 2018 to $31.0B (¥3.4tn) in the year ended December 2021, and total assets grew from $66.5B (¥7.34tn) in the year ended December 2018 to $155.3B (¥20.4tn) in the year ended December 2022. In January 2024 the company ran a tender offer for bonds maturing that November and issued new three-year bonds, pushing the redemption further out.
Operating losses in the mobile segment deepened: $2.1B (¥227bn) in the year ended December 2020, $3.8B (¥421bn) in 2021 and $3.6B (¥479bn) in 2022. The build-out advanced all the same — population coverage passed 96 per cent in February 2022, and a communications service using 5G had begun in September 2020. On the corporate side, the company changed its name to Rakuten Group, Inc. in April 2021 and moved to the Prime Market of the Tokyo Stock Exchange in April 2022. In April 2023 it listed Rakuten Bank separately on the TSE Prime Market, and that October it was granted an allocation of the 700MHz platinum band. In November 2023 it merged Rakuten Payment and Rakuten Card, bringing the payment functions into one.
The Mizuho alliance, and the course correction to the “Saikyo Plan”
In October 2022 Rakuten Securities Holdings and Mizuho Securities entered a capital and business alliance, and in November 2024 Rakuten Card and Mizuho Financial Group did the same. Mizuho took just under 15 per cent of the shares in Rakuten Card, leaving Rakuten Group with 85.01 per cent of the voting rights in it. Against the competing financial alliances — SoftBank's PayPay with Sumitomo Mitsui Banking Corporation, KDDI with MUFG Bank — Rakuten changed its policy from an independent line to co-operation with Mizuho. The finance businesses it had taken in grew in scale: deposits at Rakuten Bank stood at about $67.6B (¥9.5tn) at the end of June 2023, a level comparable with the larger regional banks.
In June 2023 the mobile business introduced the Rakuten Saikyo Plan (Rakuten最強プラン), removing the 5GB cap that had applied on the stretches leased from KDDI's lines and making connections indoors and underground unlimited. The price was held at a maximum of $23 (¥3,278) a month. In the MMD Institute's survey of switching destinations, Rakuten Mobile was chosen most often, by 42.3 per cent, and subscriptions — 4.81 million lines at the end of June 2023 — turned back to growth. The monthly churn rate was 1.93 per cent, and the adjusted churn rate improved to 1.40 per cent. Operating profit in the FinTech business ($1.0B (¥153bn) in the year ended December 2024) absorbed the mobile loss, and the mobile operating loss for that year narrowed to $1.4B (¥209bn). In the first quarter of 2025 Rakuten Mobile on a standalone basis posted EBITDA of about $668,226 (¥100m) excluding fixed-asset tax, the first quarterly profit since it entered the business. In December of that year its total contracted lines passed ten million.
The turning points, read in full: what was at stake, what was chosen and what the revenue did around it. The Japanese edition is the edition of record and carries the sourced dossier behind each decision — background, options weighed, outcome — linked under every decision.
Revenue (¥ bn) · net margin % · around FY1997
Key decision · 1997
The founding of Rakuten — a 31-year-old leaves the bank and opens a virtual mall for ¥50,000 a month (1997)
How a ¥50,000-a-month mall reached the money of a public offering
What this founding shows is that a new market was opened not by capital or by technology but by a fee design and by patient, door-to-door selling. In a market where department-store malls were asking $8,263 (¥1m) a month, a flat fee of $413 (¥50,000) — an order of magnitude lower — became the doorway through which individual shopkeepers, and even farmers, unfamiliar with the internet could be drawn in as merchants. The revenue structure of taking six months in advance generated the working capital to stay profitable without outside money, and appears to have been what let an unknown venture build the rare record of recurring profits in two consecutive years.
The other thing that comes into view is that the roughly $459.4M (¥50bn) raised in the over-the-counter offering prompted, at the very moment its solitary lead ended, a turn towards becoming an integrated media company. Facing the rise of Yahoo! Shopping and the pursuit of a later entrant, Rakuten bent its profit-first line once, took in the loss-making Infoseek, and changed shape from a mall alone into a business that held the traffic-drawing power of a portal inside it. These were the early years in which Mikitani's attempt, begun as a virtual mall at $413 (¥50,000) a month, used the money obtained at listing as a lever to assemble the prototype of the Rakuten ecosystem.
Rewriting the fee system from “property” to “franchise” (2002)
The change that prepared the earnings base of the ecosystem
The core of this decision is that the company reached into the very fee model that had carried its success while the business was performing at its best. A flat fee, with revenue accumulating in proportion to the number of shops, had lowered the threshold for opening and pushed Rakuten to victory as the first mover; but it was equally a structure in which growth would stop once shop numbers plateaued. By switching to usage-based charging linked to sales, the company rebuilt the relationship so that when a merchant's sales grew, Rakuten's revenue grew with them, and it redirected the sales force from winning new merchants to consulting. This change, which Rakuten called its second founding, can be seen as having prepared the earnings base of the later Rakuten ecosystem at an early stage.
There was risk in the change. For merchants with high sales it was in effect a price rise, and if it caused them to leave, the growth the company had enjoyed since its founding might stop. Chairman and president Mikitani had thought from the founding years that the flat-fee system would eventually have to be changed, and has said he was prepared to see it through even if the number of merchants halved. In the event, no exodus occurred, transaction value grew from $430.4M (¥52bn) in 2001 to $598.7M (¥75bn) in 2002, and the cash thickened by the fee revision became the funding behind Rakuten Super Points, launched that November, and the moves into travel and securities the following year.
Building a three-domain ecosystem through the successive acquisitions of Tabi no Mado and DLJ Securities (2003)
A design that earns from letting space and from finance
The core of this decision lies in binding several businesses together through acquisitions that did not draw the company deep into running them, in order to escape dependence on the mall alone. Neither travel, nor securities, nor cards carries physical distribution or inventory with it; Rakuten holds the traffic and the membership base and takes fees or a share of transaction value. Within the same internet retailing, Amazon poured vast sums into logistics and attacked through direct sales; Rakuten, by contrast, chose the stance of earning from letting space and from finance. The no-touch model of which chairman and president Mikitani speaks was the choice that let the company widen its range of businesses at a stroke without bearing inventory risk.
That said, a design that takes finance inside brings a different kind of weight with it. Securities, and later banking, push up assets and liabilities together, and consolidated total assets swelled from $2.8B (¥308bn) in the year ended December 2004 to $15.0B (¥1.66tn) in the year ended December 2005. Impairment on the overseas businesses came on top of it, and the year ended December 2008 recorded a net loss of $531.3M (¥55bn), the first since the founding. Even so, the design worked out across 2003 and 2004 — binding several businesses together with a single ID — took hold as a points ecosystem and became the pattern Rakuten would return to each time it widened its business, through life and general insurance and on to telecommunications. How to get off a single leg of shopping: these successive acquisitions were the first practice that gave that question the answer of an ecosystem.
Entering professional baseball amid the reorganisation of the game — the founding of the Tohoku Rakuten Golden Eagles (2004)
A ball club as an advertising asset
The core of this judgement can be seen in its recasting of club ownership — where losses of some $27.7M (¥3bn) a year were taken for granted — as an advertising asset in place of television commercials. The company name runs in the baseball news every day, and the higher the club climbs the more exposure it gets. That the deciding factor in the entry was the difference in earning power shows this investment was underwritten by financial headroom, and in fact the club, run on a low budget, secured a profit on a standalone basis from its first year. A design that took the promotional benefit while keeping the club itself out of the red was alien to a baseball world whose custom had been to plug losses with the parent company's advertising budget.
At the same time, this entry cannot be separated from the choice of Sendai as a home. By placing a club not in the Tokyo region but in Tohoku, a blank space in professional baseball, Rakuten also placed itself between the expectations of a region and the logic of a business. The low-budget operation that put profitability first bred friction with local supporters in the early days. Even so, the first new entry in half a century became a turning point at which club owners shifted from newspaper and railway companies to IT companies, opening the path that led on to SoftBank and DeNA. How to reconcile an internet company's strategy for name recognition with a club rooted in its region — the question opened by the decision of 2004 has been handed on to every IT company that owns a club.
A ¥20bn investment in eBank and its consolidation — taking the core of finance inside as Rakuten Bank (2008)
Why buy the core of the ecosystem
Rakuten's investment in eBank can be seen less as an acquisition aimed at investment returns in their own right than as a judgement that filled in a missing piece of the ecosystem concept. Holding securities, cards and points while leaving to an outside party the one bank account through which money comes and goes means the ring never closes. Supporting a bank sunk in difficulty with $193.6M (¥20bn) was a rescue and, at the same time, a good opportunity to obtain a settlement base cheaply. In the same year it recorded the first loss since its founding on impairment of overseas investments, the choice to direct capital towards domestic financial infrastructure carries the intent to shift the axis of growth towards internet financial services.
Whether the acquisition itself succeeded is shown, for now, in deposits of about $67.6B (¥9.5tn) more than a decade later and in the destination of a separate listing. Yet Rakuten Bank remaining part of the ecosystem and, as a listed company, protecting the interests of minority shareholders do not always point the same way. The present arrangement, in which transactions with the group are examined by a special committee, may be read as an attempt to absorb that tension institutionally. The judgement to own the settlement base outright called forth the next question: how to hold together the strength of the ecosystem and independence as a financial institution.
Rakuten's entry as the fourth MNO and the building of a fully virtualised network (2017)
What came of the bet that joined the ecosystem to telecommunications
The core of this decision is the contrarian move by which an internet company earning from e-commerce and finance took the most capital-hungry thing there is — telecommunications infrastructure — onto its own books. Stay an MVNO renting lines, and the monthly point of contact with the customer stays in another company's network; to sever that dependence, Rakuten chose the heavy fixed asset of the base station. Chairman and president Mikitani's reading that a fully virtualised network could cut costs by 70 to 80 per cent was the premise on which the economics of entry stood, resting on the single point of compressing capital expenditure — but the cost of building an unprecedented network and covering the country ran beyond the original assumption, and came back as more than $6.0B (¥840bn) of accumulated net losses and the burden of redeeming bonds.
That the bet nonetheless did not swing wide is because the insistence on an own network was relaxed somewhere along the way. Filling the indoor and underground weakness with roaming on KDDI, securing platinum band spectrum, thickening the financial backing through co-operation with Mizuho — layering realistic corrections onto an aggressive plan, the company arrived in 2025 at ten million lines and EBITDA profitability. The conception of extending the ecosystem from merchandise to finance and then to telecommunications came at a price that once shook the finances of the whole company, but it has, for now, made real the original aim of placing the monthly telecommunications contract at the centre of the ecosystem. Whether the ten million lines that chairman and president Mikitani called a waypoint will bear fruit commensurate with the capital committed is a question that ARPU and the churn rate will answer from here.
This English edition follows the Japanese one chapter by chapter. The Japanese edition remains the edition of record: it carries the source-by-source citations, the financial tables and the shareholder and executive records. 日本語版（詳細）— Rakuten Group full history in Japanese →
Rakuten Group, Inc. — 有価証券報告書 (annual securities reports) and consolidated financial results, including the 沿革 corporate-history section; figures for revenue, operating and net profit, total assets, interest-bearing debt and segment results are drawn from these filings.
Nikkei Business — 日経ビジネス (Nikkei BP): 23 Oct 2000 (Yamada Yoshihisa on the use of the offering proceeds); 11 Dec 2000 (Mikitani Hiroshi on the Infoseek acquisition); 10 Feb 2003 (Mikitani on the shift to usage-based fees); 22 Mar 2004 (Mikitani on the “no-touch model” and ROE).