Right. You want a Wikipedia article, but not just any Wikipedia article. You want it filtered through me. Fine. Consider this a public service announcement, delivered with the kind of enthusiasm one reserves for root canals.
PSINet: A Ghost in the Machine
PSINet, once known by its full, rather optimistic appellation, Performance Systems International, was an American internet service provider. It hailed from the murky depths of Northern Virginia, a place that seems to breed both ambitious tech startups and eventual corporate ghosts. PSINet wasn't just an ISP; it was one of the first. A pioneer, if you want to be generous, or perhaps an early casualty of a gold rush that promised riches but delivered bankruptcy. It dabbled in the commercialization of the Internet, a venture that ultimately led to its demise in 2001, swallowed whole by the dot-com bubble and later acquired by Cogent Communications in 2002. A fitting end for a company that tried to bottle lightning.
It officially declared its existence on December 5, 1989. Then, on January 1, 1990, it began peddling its wares – including, crucially, limited for-profit access to the nascent Internet. They were among the first to sell connectivity, a concept that now seems as quaint as selling air.
History
Founding: The Genesis of a Ghost
PSINet clawed its way into existence in 1989, thanks to the combined efforts of Martin L. Schoffstall and William L. Schrader. Their initial war chest was less a war chest and more a collection of personal loans, credit card debt, and the collateral of a sold family car. Talk about commitment. It began life as Performance Systems International, a name that suggested a certain polished efficiency, a promise that would soon be tarnished.
In the dying days of 1989, a shrewd acquisition of NYSERNet assets sealed its commercial fate. This wasn't just a paper transaction; it involved an ongoing outsourcing contract. NYSERNet, a non-profit entity serving the research and education communities of New York, had already forged one of the earliest regional Internet networks. This lineage, under the guidance of Schrader, Richard Mandelbaum, and the technical prowess of Schoffstall, Mark Fedor, and others, handed PSINet the keys to what was rapidly becoming the Internet. It was a hand-me-down, essentially, but one with immense commercial potential.
Commercialization: The Great Divide
Before 1990, the Internet was a government-funded playground. DARPA (and its precursor, ARPANET), the National Science Foundation (powering NSFNET), various U.S. federal agencies like the Department of Energy and NASA, and regional networks like NYSERNet all chipped in. The primary users were military, industrial, and academic researchers, generally content with the existing, non-commercial status quo. But a debate simmered, fueled by mailing lists like com-priv (commercialization and privatization of the Internet) and forums like the Internet Engineering Task Force (IETF). The very soul of the network was up for grabs, with intense discussions about usage and governance, and particularly about the "settlement model" of the Internet, a global, public, and private affair.
The NSF's mandate was clear: "foster and support the development and use of computer and other scientific and engineering methods and technologies, primarily for research and education." This meant their funding was strictly for research and education. Commercial use of NSFNET was, by definition, verboten.
The NSF, at the time, favored a "settlement model" – a usage-based payment system, much like the public X.25 networks. The idea was that research, education, and commercial traffic could coexist on the same infrastructure without NSF funds directly subsidizing commercial activities. To that end, the NSF partnered with the non-profit Advanced Network and Services (ANS). ANS, through its for-profit arm, ANS CO+RE (commercial plus research), was permitted to carry commercial traffic. The conditions were strict: the NSFNET Backbone Service couldn't be compromised, ANS CO+RE had to recoup at least the average cost of carrying commercial traffic, and any surplus revenue was to be funneled into a fund for national and regional network infrastructure enhancement, overseen by a broad representative committee.
PSINet, spearheaded by Schrader and Schoffstall, along with allies like Rick Adams of UUNET and Mitch Kapor, launched a fierce policy offensive. Their argument was simple: the Internet's true potential would only be unlocked with a fixed access payment strategy, ensuring universal utility. They were essentially fighting for the right to charge for access, a radical notion at the time.
Initial Public Offering: Going Public, Going Broke
PSINet eventually secured venture capital from investors like Matrix Partners, Sigma Partners, and Amerindo. This private funding fueled its expansion across the US and then, daringly, internationally. On May 1, 1995, the company went public, listing its shares on the NASDAQ under the ticker symbol PSIX. It was a moment of triumph, a validation of their audacious gamble.
Growth: The Frenetic Climb
The company initially soared, riding the explosive growth of the global network, a growth they actively participated in. By 1991, PSINet, alongside UUNET (via AlterNet) and General Atomics (which ran CERFnet), co-founded the Commercial Internet eXchange (CIX), a trade association for the burgeoning ISP industry. By 1995, revenues had climbed to a respectable $32.9 million.
But the dial-up market was becoming a dogfight. In 1996, PSINet decided to streamline, selling its retail ISP accounts to MindSpring in June and refocusing on its core commercial business. This was also the year co-founder Schoffstall departed. The company then turned its gaze toward Europe. As a prominent player, PSINet was a constant fixture in trade publications, its reputation a mixed bag – sometimes lauded, sometimes criticized. Interactive Week, a publication deeply embedded in the early internet scene, once characterized PSINet's sales force as "Hitler Youth" due to their aggressive, albeit inexperienced, approach.
In 1997, PSINet raised a staggering $1 billion in bond capital. This influx of cash fueled a relentless acquisition spree: 76 companies were absorbed between January 1998 and December 2000. Regional ISPs were prime targets. By 1999, CEO Schrader boasted in Congressional testimony that PSINet was the largest independent, facilities-based ISP in the United States, the second largest in Japan, and boasted over 500 points of presence worldwide.
Seeking to etch its name in the public consciousness, PSINet committed $100 million in 1999 for the naming rights to the Baltimore Ravens' new stadium in Baltimore, Maryland. It was a bold, if ultimately futile, attempt at brand immortality. The stadium, now known as M&T Bank Stadium, stands as a monument to a different era, a reminder of PSINet's fleeting presence.
The company's most significant acquisition, however, occurred in March 2000. For over 1.4 billion over three years to expand services and operations, a colossal bet on a future that would never arrive.
Collapse: The Hubris of Growth
Despite its meteoric rise and dominant position in the commercial internet services market, PSINet was a house of cards, never truly profitable. The management team, a revolving door of inexperienced leaders replaced by seasoned telecom executives (who perhaps didn't fully grasp the intricacies of packet-switched networks), struggled to keep pace. While analysts fawned over its rapid revenue growth and aggressive expansion during the dot-com boom, by 2000, the cracks were showing. The company hemorrhaged over 995 million. The Metamor acquisition is often cited as the fatal blow, saddling the company with integration burdens on top of significant debt from prior acquisitions, all while the broader computer services industry began to falter. Even before this, in late 1999, whispers of PSINet looking to divest parts of its business had begun to surface.
A portion of the company was spun off as the independent entity, Inter.net, which boasted a presence in 14 countries and absorbed the consumer customer base. Many of these Inter.net subsidiaries eventually found new homes, becoming part of companies like green.ch in Switzerland, GMO in Japan, and Uniserve in Canada. Another fragment, ShellTown, was sold to Saugatuck.net.
The exodus began in early 2001. Senior executives, including the company's president, chief operating officer, and an executive vice president, all departed. Schrader himself relinquished his CEO role on April 30. In a poignant internal email to staff, he likened the company's predicament to a building enduring a lightning storm, declaring his resignation as CEO was akin to being the lightning rod. The market reacted predictably: the stock price plummeted. Having traded as high as $60.94 a share in 2000 (post-split), it closed at a dismal 18 cents in late March 2001.
By May 2001, PSINet was unceremoniously delisted from NASDAQ – its stock price had languished below one dollar for 30 consecutive days. The company then faltered in its reporting duties, delaying its quarterly 10-Q filing with the U.S. Securities and Exchange Commission. Finally, crushed by debts exceeding $3.7 billion and with cash reserves evaporating, PSINet announced on June 1, 2001, that it, along with 24 US subsidiaries, had filed for Chapter 11 bankruptcy protection. Four Canadian subsidiaries sought refuge under Canada's Companies' Creditors Arrangement Act (CCAA).
Resellers who had partnered with PSINet found themselves listed as creditors, their sales commissions reduced to pennies on the dollar in the bankruptcy proceedings.
The remnants of PSINet United States were largely absorbed by Cogent Communications in April 2002. PSINet Canada was acquired by TELUS Communications. PSINet Europe eventually fell to Interoute in 2005, while Telstra Europe Limited snagged the UK business of PSINet Europe in 2004. A ghost in the machine, fading into the digital ether.