In the fall of 1988, an American family could sit down at a beige IBM personal computer, pick up the telephone to make sure no one was on the line, and dial into something called Prodigy. The modem screamed as it negotiated a connection at 1,200 bits per second, slow enough that each pixel seemed to arrive individually. Then the screen filled with color. Not green text on black, but actual graphics, headlines, pictures, and a banner advertisement anchored across the bottom of every page.
It was the first time most Americans had ever seen an online display ad. It was also the front door to the largest consumer experiment in the history of American telecommunications, a joint venture backed by IBM and Sears that would absorb more than a billion dollars, attract two million subscribers at its peak, and vanish from the American internet within a decade. The idea behind Prodigy had been growing since the early 1980s. In France, the government-backed Minitel system was proving that ordinary citizens would use a small terminal to look up phone numbers, book train tickets, and send messages.
American corporations wanted to do the same thing for profit. The technology was called videotex, connecting a screen in the home to a central database over a telephone line. In 1984, CBS, IBM, and Sears formally incorporated a joint venture called Trintex. IBM brought the engineering.
Sears brought the retail infrastructure and its massive catalog customer base. CBS brought the content vision. Their plan was staggering in its confidence: build a nationwide consumer information network at a time when the word online meant almost nothing to the average American family. The revenue model was radical.
Instead of charging by the minute the way CompuServe and every other service did, Trintex would charge a low flat monthly fee and make up the difference with advertising revenue from corporate partners. It was, in everything but name, the business model of the web, conceived a decade before the web existed. The coalition was unstable from the start. In 1986, CBS chairman Thomas Wyman was forced out in a boardroom struggle, and his successor began selling off non-broadcast assets.
CBS withdrew from the partnership. IBM and Sears pressed on alone. By the fall of 1988, they had invested $450 million in development. They renamed the venture Prodigy, a name chosen by a branding firm that carried the weight of a promise: this service would be ahead of its time.
The technical foundation was impressive. Engineers at IBM built a network of minicomputers deployed in regional hubs across the country. Each hub stored the most frequently accessed content locally, so a subscriber in Atlanta dialed across town rather than across the country. It was an early content delivery network, built years before the term existed.
On the user side, the interface was rendered using vector graphics and a proprietary programming language. The screens were bold, bright, and deliberately simple. One technology columnist who connected in 1988 described the interface as pretty grim, simplified to the point of near imbecility. But the simplicity was the point.
Prodigy was not designed for hobbyists who could navigate a command line. It was designed for the family that had just bought its first PC and never used a modem before. To get that family connected, Prodigy partnered with Hayes Microcomputer Products, the company that had essentially created the consumer modem market. Hayes designed a 1,200 baud modem exclusively for Prodigy subscribers, packaged it with software, floppy discs, and three months of free service, and sold the whole startup kit for $149.
95. To fill the screens with something worth reading, Prodigy hired Jim Bellows, a legend in American journalism who had edited the New York Herald Tribune and helped launch ABC’s news operation. Bellows assembled a newsroom of reporters, writers, editors, and graphic artists producing original content for a screen instead of a printed page. It was the first sustained experiment in digital journalism in American history, launched years before any newspaper put a single word on the internet.
Prodigy launched commercially in September 1988 in Atlanta, Hartford, and San Francisco. By mid-1989, only 65,000 households had signed up, far below expectations. The modem was still a strange device to most families, and the two parent companies, already half a billion dollars in, were beginning to wonder whether anyone wanted what they were selling. The answer came on September 6, 1990, when Prodigy went national.
A television advertising campaign put the service in front of the mass market for the first time, and the flat-rate pricing did the rest. $9. 95 a month, all you could use, no long-distance charges. At a time when CompuServe charged $6 to $12 an hour, the math was not even close.
Nearly 100 advertisers signed on: Sears, American Airlines, JC Penney, Dean Witter, Levi Strauss, Ford, Buick, Volkswagen, Procter & Gamble. Every screen carried a banner advertisement across the bottom, the direct ancestor of every banner ad that later colonized the web. By the end of 1990, Prodigy had 635,000 subscribers. By October 1992, the number reached 1.
75 million. At its peak, it claimed two million members, making it the largest online service in America. But inside that growth was the seed of collapse. The flat-rate model meant heavy users consumed enormous amounts of bandwidth at no additional cost to themselves.
A small fraction of members generated millions of email messages, routed through regional points of presence that required expensive leased lines to keep synchronized. Advertising revenue was not growing fast enough to cover the cost of serving an audience that used the service far more intensely than anyone had projected. The model that attracted two million subscribers was bleeding the company dry. At its height, the White Plains headquarters employed a thousand people building and maintaining a service that connected two million American families to a shared digital world.
It offered more than a hundred content areas: news, weather, sports, stock quotes, airline reservations, home banking, an online encyclopedia, games, and electronic mail. For a few years, it was the closest thing the United States had to a national information utility. It had products that people loved. Mad Maze, a text-and-graphics adventure game developed in 1989, became the first online game to draw more than a million players.
The bulletin boards, moderated by volunteer experts called board leaders, were used with an intensity that surprised even Prodigy’s management. The most popular was Money Talk, a financial discussion forum where subscribers debated stocks and exchanged investment tips. It was one of the most active financial communities in the country and a direct ancestor of every modern stock forum. The company’s management then began dismantling the community it had built.
In 1991, Prodigy began censoring its bulletin boards with a heavy hand. Messages criticizing advertisers were deleted. Profanity was blocked so aggressively that a subscriber could not use the word in a dog breeders forum. Users who organized protests against the moderation policies were disconnected from the service entirely.
In November 1990, the Washington Post reported that dissident users had been cut off for challenging the company’s content rules. Users responded by forming underground email networks to circulate the censored discussions, and the traffic overwhelmed the system. Prodigy’s answer was to cap free email at 30 messages per month, then charge 25 cents for each additional message. In 1993, it introduced hourly charges for its most popular features, including the bulletin boards themselves.
Thousands of subscribers canceled. The monthly fee climbed from $9. 95 to $12. 95, then to $14.
95. Each increase pushed more people toward the exits, and the people leaving were not going offline. They were going to America Online. AOL had started as a dial-up service for Commodore 64 users before rebranding in 1989, the same year Prodigy adopted its new name.
But while Prodigy was managed like a corporate product, AOL was run by founder Steve Case with the instincts of a startup. Case understood something IBM and Sears did not. People did not go online primarily to shop or read the news. They went online to find other people.
AOL offered chat rooms, instant messaging, anonymous screen names, and member-created communities. It marketed with ferocity, flooding the country with free trial discs until it was a running joke that you could not open your mailbox without an AOL floppy falling out. On Prodigy, you landed on a curated highlights page that looked like the front page of a carefully edited newsletter anchored by an approved advertisement. On AOL, within two clicks you were in a chat room full of strangers typing things no IBM content policy would have allowed.
The American public, given the choice between the curated and the chaotic, chose the chaos. Then came the open internet. In 1993, a graduate student at the University of Illinois released Mosaic, the first web browser with a graphical interface, and the internet stopped being an academic network and became a consumer revolution. The open internet offered everything Prodigy offered and more, without a gatekeeper, without a sensor, without a banner ad hard-coded into every page.
AOL pivoted toward internet access and positioned itself as the friendliest on-ramp to the web. Prodigy, managed by executives trained in the cultures of IBM and Sears, pivoted more slowly. The institutional reflex was to control, curate, and moderate, and the open internet was none of those things. Then came the lawsuit.
In October 1994, an anonymous user posted a message on Prodigy’s Money Talk bulletin board accusing the Long Island brokerage firm Stratton Oakmont and its president of criminal fraud in connection with a stock offering. The allegations, as it later emerged, were substantially true. But Stratton Oakmont sued Prodigy for defamation, arguing that because Prodigy moderated its bulletin boards, employed board leaders, and enforced content guidelines, it was functioning as a publisher and should be liable for every word its users posted. The New York Supreme Court agreed.
In 1995, the court ruled that Prodigy’s conscious choice to exercise editorial control made it a publisher legally responsible for the statements of its users. The ruling created a perverse incentive. Any service that tried to moderate content could be held liable for all of it, while a service that moderated nothing escaped liability as a mere distributor. The message to every company in the business was brutal: if you want to avoid lawsuits, stop trying to keep your platform decent.
Representative Chris Cox, a Republican from California, read about the decision in the Wall Street Journal and contacted Representative Ron Wyden, a Democrat from Oregon. Together they drafted legislation. The House passed it 420 to 4 in August 1995. It was incorporated into the Communications Decency Act, which the Senate approved 91 to 5 in February 1996.
When the Supreme Court struck down most of the law the following year on First Amendment grounds, one provision survived. Section 230 declared that no provider of an interactive computer service shall be treated as the publisher or speaker of information provided by another user. Section 230 is still the law. It is the legal foundation on which Facebook, YouTube, X, and every platform that hosts user-generated content operates today.
It exists because Prodigy tried to keep its bulletin boards family-friendly, was punished for it in court, and prompted two congressmen to fix what the ruling had broken. By 1996, the numbers were merciless. AOL had six million subscribers. CompuServe had five million.
Prodigy had fallen to 1. 3 million, and the trajectory pointed only downward. In three years, it had gone from the largest online service in America to a distant third. In May 1996, IBM and Sears gave up.
They sold Prodigy for roughly $200 million. They had invested at least $1. 2 billion. A company built on the promise of connecting America to the future was sold for less than a fifth of what it cost to build.
The new owners tried to reinvent what remained. They split the service and took the company public in February 1999. The stock climbed to $50 in the fever of the dot-com bubble, and for a handful of months the company was valued at over a billion dollars on paper. The paper did not hold.
On October 1, 1999, Prodigy Classic was shut down. The official reason was that the aging code base could not be made Y2K compliant at a cost anyone was willing to pay. More than 200,000 subscribers remained on the old service, down from two million at the peak. The servers, the regional points of presence, the bulletin boards where a million people had argued about money and politics and life, all of it went dark on a single day.
In November 1999, SBC Communications took a 43% stake in what remained, pledging to route broadband customers through Prodigy’s infrastructure. In 2001, SBC bought the rest. In 2005, SBC acquired AT&T itself and took its name. Prodigy’s subscribers, its network, and whatever remained of its brand were folded into the largest telecommunications company in the United States.
The name vanished from the American internet entirely. But not from Mexico. Carlos Slim’s Telmex had taken Prodigy south, and under the brand Prodigy Infinitum, it became the dominant broadband provider in that country, commanding an estimated 92% of the market at its peak. The service that could not survive the country that built it conquered the country that adopted it.
The legacy survived elsewhere too. Section 230 still governs how every social media platform on earth handles the words of its users. The advertising model Prodigy pioneered, selling banner ads against user attention on a screen, now generates hundreds of billions of dollars a year. Online shopping, online banking, online news, and online gaming are so deeply embedded in daily life that the world without them requires an active imagination.
The company that proved all of these ideas could reach a mass audience did not survive long enough to profit from any of them. In 2014, a software developer acquired old hard drives from the early 1990s and discovered a cache file that Prodigy had used to store downloaded page data locally. When a subscriber connected for the last time, the data in that file froze. Every headline, every advertisement, every game menu, every blinking mail icon was preserved inside the file like an insect in amber, untouched for more than twenty years.
He wrote scripts to extract the data, decoded the graphics, and began pulling Prodigy back from the dead. A technology journalist documented the work and founded the Prodigy Preservation Project. People dug old computers out of attics and donated floppy discs they had kept for decades without knowing why. In 2022, at the Vintage Computer Festival West in Mountain View, California, a programmer demonstrated a working reconstruction of the Prodigy service that could accept connections from original 1990s client software, complete with login, email, stock quotes, and Mad Maze.
For the first time in 23 years, a user could dial into Prodigy and see what two million American families once saw when the modem finished screaming and the screen filled with color. IBM and Sears spent more than a billion dollars building the online world before the online world existed. They hired a legendary newspaper editor to create the first digital newsroom. They shipped the first banner advertisement on a consumer screen.
They hosted the first online game to attract a million players. They connected two million homes to a network and charged $9. 95 a month for the privilege of living in the future. Then the open internet arrived and offered all of it for free.
Prodigy did not fail to imagine the online future. Prodigy built it, charged admission, and watched the open internet hand it to everyone for nothing.