Internet Initiative Japan: long-term performance & turning pointsSales (revenue) and profit-margin ratio
Sales (¥ bn)Net margin (%)
1992Japan’s first commercial ISP — and its first crisis
Revenue (¥ bn, bars) · net margin (%, line)
Source: securities reports
1992Founded in Tokyo as Internet Initiative Japan Planning
1993Commercial internet-connection service begins
1994Registered as a special Type II telecom carrier
1996US subsidiary IIJ America established
1999ADRs listed on Nasdaq
2003Crosswave collapses; $103.5M (¥12bn) from NTT — becomes an NTT affiliate
IIJ began in December 1992, when Koichi Suzuki set up the company in central Tokyo with capital of $142,102 (¥18m). At that moment no operator in Japan sold internet access commercially at all: the only thing running was the non-profit interconnection of research institutions (JUNET / WIDE). Suzuki carried to the banks a business plan that projected “tens of millions of users by around 2000,” and, as he later told it, they would not take him seriously. He pressed on anyway — renaming the firm Internet Initiative Japan in May 1993, launching a commercial internet-connection service that July, and in February 1994 registering with the Ministry of Posts and Telecommunications as a special Type II telecommunications carrier. Converting an internet built for researchers into a priced product meant, in effect, getting the regulator to open its rules to a commercial ISP for the first time.
Through the late 1990s IIJ built out the infrastructure of a network company. It set up the US subsidiary IIJ America in 1996 as the anchor of a Japan–US backbone, formed Internet Multifeed with NTT in 1997 to run an interconnection point (IX), and in October 1998 launched Crosswave Communications as a joint venture to carry a telecom-carrier business. In August 1999 it registered ADRs on the US Nasdaq market — an early Nasdaq listing for a Japanese company that opened a funding route to American institutional investors.
The carrier venture became the company’s worst wound. In August 2003 Crosswave — into which IIJ had poured capital during the dot-com boom’s over-investment in carrier infrastructure — filed for court-led reorganization, and the loss flowed through to IIJ’s books. The following month IIJ raised $103.5M (¥12bn) in a third-party allotment underwritten mainly by NTT, becoming an NTT equity-method affiliate and, with that, surrendering its standing as an independent ISP. The lesson was blunt: a capital-intensive, slow-to-recover carrier business was more than IIJ’s own balance sheet could bear. Suzuki would later say of it, “looking back, it was not even half-done — it was eighty per cent regret.”
2008Japan’s first MVNO enterprise mobile-data service
2009IIJ GIO cloud service launches
2012IIJmio — Japan’s budget-SIM market opens
2013Eijiro Katsu becomes president; Suzuki to chairman & CEO
2018Japan’s first full MVNO
IIJ listed on the Tokyo Stock Exchange’s Mothers market in December 2005 and, barely a year later, moved up to the First Section — clearing the carried-forward losses from the Crosswave collapse along the way. But the more consequential change was in how it earned. Chastened by an investment whose payback it could not control, IIJ set about redesigning its revenue itself: away from flow-type ISP connection and toward an enterprise stock model, in which monthly contracts accumulate month after month.
That turn ran through a string of firsts. In January 2008 IIJ began offering enterprise mobile data on a wholesale line from NTT Docomo — Japan’s first MVNO. In December 2009 it launched the IIJ GIO cloud service, among the first commercial public clouds in Japan. In 2010 it absorbed the domestic network-outsourcing business of AT&T Japan as IIJ Global Solutions, taking a foreign carrier’s Japanese operations and, with them, a foothold in the corporate-network market.
The build-out of infrastructure and consumer reach followed. In April 2011 IIJ opened the Matsue Data Center Park in Shimane — an early free-cooling, container-type data centre in Japan. In February 2012 it launched IIJmio, whose cheap SIM-based mobile data created Japan’s budget-SIM market and pushed the MVNO business from B2B into B2C. In June 2013 Suzuki moved up to chairman and CEO, and Eijiro Katsu — a former administrative vice-minister of finance — became president, IIJ’s first top executive recruited from outside. In 2018 IIJ launched Japan’s first full MVNO, owning even the SIM issuance and management functions.
2019Delists ADRs from Nasdaq; opens Shirai Data Center Campus
2022Moves to the TSE Prime market
2023Leaves NTT’s affiliation; capital-and-business alliance with KDDI
2025Yasuhiko Taniwaki becomes president — a “second founding”
By the end of the 2010s the stock model had become an engine of steady, structural profit growth. In 2019 IIJ delisted its ADRs from Nasdaq — roughly twenty years after listing — and opened the Shirai Data Center Campus in Chiba as a wholly owned core asset for its cloud and operations business, later a base for sovereign-cloud demand. Enterprise recurring revenue and step-by-step gains in operating profit took hold: in the year to March 2021, operating profit jumped by more than seventy per cent as remote-work, DX and cloud demand lifted IIJ’s main services all at once.
In May 2023 came the reversal of 2003. NTT sold down its stake, IIJ left NTT’s equity-method affiliation, and it simultaneously signed a capital-and-business alliance with KDDI — leaving it a top shareholder shared equally between the two telecom giants. Twenty years after the Crosswave rescue had forced it under NTT, IIJ had rebuilt a capital structure equidistant from both, restoring — in ownership terms this time — the independent-ISP concept it was founded on. The alliance also strengthened its hand as a full MVNO, letting it secure wholesale lines from both NTT Docomo and KDDI.
The generational handover followed the same logic of recruiting expertise from the state. In April 2025 Katsu retired and Yasuhiko Taniwaki — a former director-general of the telecommunications bureau at the Ministry of Internal Affairs and Communications — became president, calling this IIJ’s “second founding” and pointing the company at a data-distribution business and at government sovereign-cloud demand. In February 2026 IIJ set up Sensifia, an IoT joint venture with Sony Semiconductor Solutions. Revenue reached $2.1B (¥317bn) in the year to March 2025 — roughly eight times its scale at the time of the Crosswave collapse — and chairman Suzuki now speaks of “a company that can see ¥1 trillion in sales.”
The turning points, read in full: what was at stake, what was chosen and what the revenue did around it. The complete sourced record of each — background, options weighed, outcome — is in the Japanese edition, linked under every decision.
Key decision · 1992
Founding IIJ in a country with no commercial ISP (1992)
Who is first to open demand that has no price yet
The core of this venture lies less in the novelty of the technology than in being first to put a price on demand no one had yet priced, and to open it as a market. To replace connectivity that flowed free among researchers with a product enterprises paid to use, there were no numbers to prove the demand was real and no precedent to point to. The business plan the banks dismissed as a “tall tale” was, turned the other way round, evidence that this demand was still invisible to everyone. A judgment to bet on demand that could not be seen is what opened the door to Japan’s later internet industry.
Being first, however, did not mean being safe. Amid the dot-com boom’s euphoria and its backlash, IIJ lost its capital independence once, and only after a long labour of rebuilding its business toward a stock model did it recover stable profit growth and its independence together. The one who first opens a market is not guaranteed to remain its protagonist. IIJ’s course reflects both the advantage of going first and the trial that going first imposes. Who is first to open demand that has no price? That question, in changed form, keeps reappearing in the contest over the next digital infrastructure.
The Crosswave collapse and taking capital from the NTT group (2003)
What, exactly, was independence meant to protect
The character of this capital move cannot be grasped as merely the financial mopping-up after a failed business. IIJ, which through the Crosswave joint venture had challenged the telecom infrastructure that was NTT’s own domain, took capital from that very NTT once the challenge collapsed. In the gap between the banner of defiance that an independent-ISP standard-bearer had raised and the outcome of entering the largest carrier’s capital keiretsu, one glimpses the weight of carrying a capital-intensive business on independent capital alone. When Koichi Suzuki later said it was “not even half-done — eighty per cent regret,” it was, it seems, this gap he meant.
Seen another way, the episode forced IIJ to rethink how it earned in the first place. Having learned the weight of investment that acquires plant up front and then waits to recover it, the company shifted its mainstay to an enterprise stock model of accumulating contracts, and twenty years later stepped out from under NTT’s umbrella into a capital structure equidistant from KDDI — a result of weaving the lesson of the collapse into both business design and capital policy. What does independence point to — independence of capital, or the independence of standing a business up on one’s own strength? IIJ’s twenty years mirror that question from both the capital structure and the earnings structure at once.
From flow-type ISP connection to the enterprise stock model (2012)
Choosing to change the way of earning, not the capital spend
The heart of this decision lay less in which businesses to enter than in reworking how revenue was accumulated. The Crosswave collapse had confronted IIJ with what happens when a large investment whose recovery cannot be read outruns a company’s own strength. Rather than retreat or shrink, IIJ chose to absorb that lesson by redesigning the very way it earned — toward a stock model in which monthly contracts pile up. The distinctive thing about this turn is that it was a redesign of the way of earning, not a one-off round of capital spending.
That said, MVNO, cloud and data centres are all businesses heavy with up-front investment and price competition; recasting them as stock does not by itself buy safety. Even so, a revenue base built on continuous enterprise billing underpinned the structural profit growth of the 2020s and the recovery, after twenty years, of capital independence. From a flow-type connection specialist to a company that stacks up contracts to support the recovery of its plant — how to judge this turn depends on whether the next stage of growth, flying the flags of government demand and data distribution, can be carried on the same shape of revenue.
Leaving the NTT group and allying with KDDI (2023)
From receiver of capital to designer of it
The core of this decision can be seen in how IIJ’s position shifted — from a receiver of capital to a designer of it. Whereas the 2003 capital acceptance was an unavoidable choice forced by a collapse, the 2023 reshaping was a company that had recovered its financial autonomy choosing, for itself, a capital structure that leaned toward no camp. The path of reclaiming, in order, first the independence of the business and then the independence of capital converged here into one.
Even so, how far a design equidistant from NTT and KDDI translates into strength as a neutral operator depends on the substance of the collaboration to come. A structure holding two equal top shareholders makes it easy to widen dealings with both, yet carries a balance in which it is hard to step deeply toward either. For a company placing government demand and sovereign cloud at the centre of its next growth, how it turns a capital structure that faces the two giants equally to advantage is what will be asked of it from here.