← all stories

Before the Web · Era 1 · Boards, networks and online services

1984–2001sold in 1996 for $78.1 million (confirmed by a primary…

Prodigy

Prodigy was a closed online service with advertising on the screen and a flat rate, started jointly by IBM, Sears, and CBS in 1984 and launched in 1988–1990. It was the first mass-market service to show graphics and a family internet-before-the-internet, but it lost to America Online on the speed of its development and landed in a court case about censorship of its forums that gave rise to the famous Section 230, and in 1996 it was sold at an enormous loss to its owners.

Founders IBM · Sears, Roebuck & Co. · CBS (left in 1986) · Theodore Papes (CEO of Trintex/Prodigy, 1984–1992)
Domains prodigy.com · prodigy.net
pre-internetad-supportedclosed-platformplatform-liabilitycorporate-jvcontent-moderation

In brief

Prodigy was a closed online service of the late 1980s and early 1990s: a flat rate of $9.95 a month, advertising on every screen, graphics instead of a command line. It was started in 1984 under the name Trintex by three giants, IBM, Sears, and CBS (CBS left in 1986), and shown to the public in 1988–1990. At its peak in 1994 Prodigy had about two million subscribers, but it was overtaken both by CompuServe and by the fast-growing America Online. Prodigy is remembered less for its success than for the lawsuit brought by the brokerage Stratton Oakmont (the one from The Wolf of Wall Street): it was punished for exactly what others were getting away with, for an attempt to moderate its forums. That case gave rise directly to Section 230, the law that to this day determines who answers for other people's words on the internet. Prodigy was sold in 1996 for $78.1 million, a sum the press of the day was unable to name, against accumulated operating losses of $1.3 billion at its owners.

How it started (the founders)

Prodigy has no story of garage founders. On February 13, 1984 IBM, Sears, and CBS registered the joint venture Trintex; the idea had grown out of the videotex experiments of the early 1980s, in which CBS, together with AT&T, had already tested a similar service on 100 homes in New Jersey. Theodore Papes, a career IBM manager, was put in charge, and stayed until his resignation in 1992. Behind it stood three conglomerates that put in hundreds of millions of dollars before the first subscription: by the fall of 1988 the start-up costs came to $450 million.

In 1986 CBS left the project: the new CEO, Laurence Tisch, was selling off non-core assets, leaving IBM and Sears as equal partners. Several more years went into developing a graphical platform of their own, the NAPLPS standard and the TBOL and PAL languages, before the service saw the light of day under the name Prodigy.

Year-by-year timeline

Lesser-known but significant facts

  1. Trintex had a prehistory back in the early 1980s: CBS, together with AT&T, tested a similar videotex service on 100 homes in New Jersey, several years before CBS went into Trintex.
  2. Prodigy's graphics were ahead of the era: images were transmitted as vector commands (NAPLPS), and interpreted code was loaded onto the disk and executed locally, which in hindsight gets compared to the later Java applets estimate. IBM's patents on these ideas are real: in 2018 a jury awarded IBM $82.5 million from Groupon for infringing them, but the patents are not still in force, since all four had expired by 2025.
  3. The columnist Jerry Pournelle criticized the service at a presentation in 1988, calling the interface simplified almost to the point of idiocy and lacking a mouse, and worked out that for $119.40 a year (exactly $9.95×12) one would do better to subscribe to a newspaper and the Wall Street Journal.
  4. The rising prices gave birth to the trick of underground IDs: several people shared an account and corresponded free of charge through deliberately wrong addresses, so that a message came back to the sender and the next person read the bounce as the message.
  5. The suit was brought by the firm of the co-owners Jordan Belfort and Daniel Porush. Belfort and Porush were later convicted of systematic fraud, not over the offering the anonymous poster of 1994 wrote about, but the characterization of the firm as criminal was borne out by the sentences of 1999–2003 and by the firm's expulsion from the NASD.

Legend vs. the record

The first growth lever

The lever was not advertising in itself but a channel of distribution that the competitors physically could not use: Sears sold Prodigy in its own stores, and IBM preinstalled it on the PS/1 and PS/2. The starter kit at $149.95 (modem, software, three months free) was sold as an ordinary consumer good. Neither CompuServe nor AOL could bring its own retail or computer manufacturing to bear.

The second element was the graphics: the NAPLPS interface with menus and pictures looked simpler and friendlier than the text command line of the CompuServe of that era, which, together with the positioning as something for the whole family, opened the service to a non-technical audience. Advertising on every screen held the subscription at $9.95 a month, against competitors oriented toward hourly billing.

Even with a channel like that, the ramp-up was slow: by the middle of 1989, 9 million homes could in theory connect Prodigy, and only 65,000 subscribed. The push came from the nationwide campaign for the launch of 1990 (the agency J. Walter Thompson), and the base grew to 465,000 subscribers by launch day.

Parallels today (projects from the catalog)

What a builder can take from this in 2026

  1. If a product rests on other people's content, choose explicitly: either you moderate fully and carry a publisher's liability, or you do not touch moderation at all. The pit is in the intermediate position of moderating a little and answering for nothing. The same choice faces AI platforms and UGC aggregators in 2026.
  2. Retail or the distribution of a conglomerate behind your back is a one-off, non-reproducible advantage rather than a strategy for a startup: it worked for Prodigy because Sears and IBM literally owned the channel. It is worth looking not for somebody else's channel but for your own, a community or a partner who cannot repeat the distribution alone.
  3. Do not touch a working rate for the sake of extra revenue if your audience is loyal but sensitive to price: both of Prodigy's attempts to earn more (1990, 1993) rolled back after an open revolt.
  4. Public words about a company's values are a potential legal document: Prodigy lost its case precisely because it called itself a "family" publication with editorial control. And the death of a company is not always binary: the Prodigy brand survived 1996 and 2001 and went on working in Mexico.

Discrepancies and what we could not verify

Sources (primary first)

Secondary:

The same thing, about today

The same breakdowns, but of projects launching right now: what the product is, where the first users came from, how they charge. The card is free, the full dossier is $5 (the dossier itself is written in Russian).

Browse the dossier catalog →