(FORTUNE Magazine) – A LOT OF COMPANIES with high hopes and deep pockets have lost money on videotext services over the years, and the few successes have been modest at best. The idea behind videotext is intriguing: Provide consumers with easy electronic ways to get information, buy goods and services, and pay their bills. Now two companies with pockets that reach to their ankles are in the game -- IBM (1986 sales: $51.3 billion) and Sears ($44.3 billion). They have already sunk an estimated $250 million in a joint venture they call Trintex, and stand to throw in millions more -- with no great demand yet in sight. Success would mark these managers as visionaries on Wall Street, of course, but other giants have cut and run from videotext investments. CBS, an early partner in Trintex, was unwilling to wait for the pot of gold IBM and Sears expect to find in the next decade, and pulled out after dropping $40 million. In 1983, after spending some $25 million experimenting with a teletext system using home TV screens, Time Inc., FORTUNE's publisher, decided not to proceed, and in 1986 it withdrew from Covidia, in which it was a 5% partner in a joint venture with AT&T and Chemical Bank to deliver banking services and entertainment listings. What draws IBM and Sears on is the same old dream that other videotext pioneers found so enticing and costly, of a world in which computer-based communications systems will be as ubiquitous as today's telephone. To turn dream into reality, hundreds of programmers and technicians have labored monklike for nearly three years in a nondescript office building in White Plains, New York, developing a product that is still about a year from coming on the market. (The product is so secret that FORTUNE was not allowed to photograph the monks at work.) The sorry history of the business does not discourage Theodore C. Papes Jr., a former boss of IBM's European operations who is CEO of Trintex. Says Papes: ''We're creating something so compelling that you'll say, 'I've got to have it.' '' The product in this case is called Prodigy, described by Trintex as an ''interactive personal service'' and generally known as a public-access videotext (PAV) network. It's an on-line computer service -- that is, you call it up -- that lets you both send and receive data over a telephone line, with the aid of a modem. Hartford will get Prodigy sometime next year, with Atlanta, San Francisco, and Boston most likely to follow. The company will charge about $10 to $15 a month for the service, roughly the price of cable television. Trintext hopes to have a nationwide system in place in three years. Prodigy expects to collect most of its revenue by sending to users commercial messages from 60-odd advertisers such as GMAC and Levi Strauss, which have signed up with Trintex for its initial run. Advertisers will pay fees based on the number of customers who actually receive a message on their screens. That degree of accuracy in measuring exposure to an ad has previously been unattainable. Prodigy will also allow advertisers to target users: for example, female runners under 35 who buy champagne. The system Trintex is building is huge -- designed to handle ten million subscribers -- and expensive. With about 600 full-time employees plus a host of loaners from IBM, Trintex can succeed only by making Prodigy a mass-market whiz in what has been a niche business. Other on-line companies such as the Source, CompuServe, QuantumLink, and Dow Jones News/Retrieval sell their services to one or both of two kinds of consumers: computer nerds who love to play games, write programs, and talk to fellow nerds, and information junkies like journalists and investors who can't bear to be out of touch with the greater world. All together these services have about a million users. Prodigy is an extraordinary attempt to prove that the theory behind interactive systems is sound and that only proper execution was lacking during the adolescence of personal computers in the early 1980s. Lots of smart people were predicting then that the PC would become the new focal point for all kinds of transactions and information. The technology was available, a market of sorts existed, but, as analysts say, the systems lacked value-added utility. In other words, videotext didn't do squat for most people. TO TRY to add value, Trintex will offer a jazzy graphics package and an easy-to-operate -- okay, user-friendly -- program that doesn't force the viewer to run through enough commands to set up a moon shot when he just wants the weather map. Prodigy is designed to use only five to eight command words displayed along the bottom. ''We drag people in off the street, give them 30 seconds of instructions and see if they can get going on the system, and generally they can,'' says Harry E. Smith, senior vice president of commercial marketing. One of the commands, called Path, lets the operator cruise along a previously selected list of service items. Subscribers can elect to get the news first, the grocery store next, their bank account third, and so on, by making just one keystroke per step. There is no need to go to a master menu each time and then through another series of commands just to view a single section. This ''personalization'' by the user is what Trintex executives believe will make the system attractive. Subscribers will need powerful machines to use Prodigy because it sucks up enormous amounts of computer capacity: To produce a full-screen image of near- photographic quality can take more than 64,000 (64K) bits of information -- the entire memory of many lower-priced computers. Initially a subscriber will have to own an IBM or compatible PC that can store 256,000 (256K) bits of information in its memory. The user will also need a modem, a color monitor, and a graphics card -- an electronic device that allows the monitor to display illustrations. Right now that relatively high-powered setup requires at least a $1,200 investment. THE PLANK to which Trintex is nailing its business plan is the same one that supported the hopes of previous failures. With other pioneers, IBM and Sears believe that the number of homes giving shelter to a personal computer will increase exponentially as the price of the technology continues to slide. Today more than ten million U.S. households have one. But most of these machines are not powerful enough to use Prodigy, and fewer still have a modem. To break even, Prodigy will need several million subscribers. The company believes that it will have a chance to get them when the price of computer, modem, and color monitor is under $1,000, and that a true mass market will develop when the price drops below $500. Trintex isn't saying when it expects to see the $500 average, but CompuServe, another on-line network, forecasts 13 million homes will have computers with modems by the end of 1992. Even if they do, will the occupants want a shopping and information service? Experience indicates most will not. Says Martin Nisenholtz, director of interactive media for Ogilvy & Mather: ''Videotext was born of technology, not of any particular need. The challenge is, and has been, to provide a service that is attractive to consumers. It is as simple as that.'' The predicament goes back to videotext's earliest days. The pioneer Qube, a joint venture of American Express and Warner Communications, got gobs of publicity but few users. The problem persists. American Airlines recently turned down an invitation to invest in Trintex, though Prodigy customers will be able to use American's Eaasy Sabre reservation system. Eaasy Sabre has been available to consumers on several videotext services for the past 18 months. The results have been unimpressive, says Max Hopper, the airline's vice president of information systems. Even consumers who have access to Eaasy Sabre find it more convenient to pick up the phone and call the airline the old-fashioned way. Computer owners have not proved to be prodigious shoppers. Despite catalogues and television shopping networks, about 90% of all retail sales are still made in stores, according to Touche Ross. The accounting firm projects that nonstore sales will account for no more than 20% of the total in a decade. Gateway, owned by Times Mirror, offered an electronic shopping service that ''went to hell in a handbasket,'' recalls James Holly, who ran Gateway. ''Our customers just had no interest in general shopping.'' CompuServe says its Electronic Mall service is successful, as does CompuCard, another shop-by- computer system, but neither service has many subscribers. Nancy Beckman at the Source puts it succinctly: ''We suspect Trintex is making a big mistake putting too much emphasis on the shopping.''

The things computer users seem to like to do are play games and talk to other users. QuantumLink, a small but profitable on-line service for Commodore computer owners, reports that its subscribers want hobby-type activities. They like games, talking with one another through a service called People Connection, and exchanging software programs. Gateway also found that its subscribers, another convivial group, liked to chat. Prodigy's hardware and software profile was forged by a culturally skewed intellectual battle among the troops who produced it. ''It was political warfare from day one,'' says an early witness. Sears, America's big store, had been selling goods through a remote channel, its catalogue, for a century. In one sense Trintex was merely an electronic catalogue. Looking deeper, though, the company could foresee retail patterns changing under pressure from time- and information-short consumers, and it also had other things to hawk: insurance from Allstate, financial services from Dean Witter, and real estate from Coldwell Banker. From the outset the Sears side took the position that the product should embrace as many consumers as possible, with the goal of making shopping easier. One participant explains: ''The customer is king, and he who owns the customer is prince. If I own the customers, I can migrate them to a greater technical ability.'' The Sears approach suggested low-end computers or even, gasp, something else, like the television-phone link called Telaction that J.C. Penney is tinkering with. And then there was Big Blue. The IBM gang pressed for more serious hardware. Says Chuck Kushell, a former Trintex marketing director who is now with Hill Holliday, an advertising agency in Boston: ''It's always been PC, PC, PC. The joke was that it was going to have to be a water-cooled PC they were trying to push so many mips (millions of instructions per second) into it.'' Papes, however, argues that Trintex has designed a system that could be adapted if necessary to work with a coming generation of ''smart'' television sets, as well as with computers. Says he: ''We are not painted into a corner. We can put our system in any box you want.'' By the time Prodigy is up and running, it probably will face plenty of competitors. A firm called U.S. Videotel has bought rights to the Minitel terminal, which the state-owned French telephone system has put into 2.5 million homes since 1981 -- where most of them are unused and collecting dust. U.S Videotel plans to launch a Minitel-like system this month in Houston. J.C. Penney's Telaction will be primarily a shopping service. Customers view products by calling Telaction on the phone and turning their televisions to a specified cable channel. By touching the digits on their phones they can stroll through the catalogue as pictures of the items appear on their home screen along with descriptions and prices. Penney has invested about $40 million in the system, which is being test-marketed in Chicago. Trintex also has to contend with the possibility that the regional Bell operating companies (RBOCs) could enter the business. The RBOCs are enjoined from doing so before 1991 by the federal court decision that broke up AT&T. But a House subcommittee had a hearing in July on whether the RBOCs' inability to get into the information race is causing the U.S. to fall behind other nations in home computerization. Trintex and other information services fear that the RBOCs could gain a significant competitive advantage, since they bring both the bats and balls to the game. Even if the RBOCs don't get into the business themselves, the Federal Communications Commission may allow them to impose added fees on the transmission of data by telephone, on the theory that data transmitters tie up lines for long periods and should pay the full cost of the service they receive. In the industry, Prodigy's naysayers are vocal. ''The longer I was there, the more I was convinced they weren't on to something,'' says Kushell, the former Trintex marketing man. But Prodigy has some surprising rooters. The on- line companies like QuantumLink and CompuServe, while fearing the prospect of a big competitor, hope Trintex will make consumers more aware of videotext -- and more willing to spend money on it. ''Trintex is going to advertise like crazy to build the market, and that will help all of us,'' argues Stephen Case, QuantumLink's vice president of marketing. That's called ''legitimizing a marketplace,'' something for which IBM happens to be famous. Whether it can legitimize this particular offspring remains to be seen.

CHART: NOT AVAILABLE CREDIT: ILLUSTRATION BY JAVIER ROMERO CAPTION: VIDEOTEXT MARCHES ON DESCRIPTION: Information about videotext companies.