Interactive Television: When's the Show Start?

by Joe Flower


A version of this article was published in March 1995 in New Scientist.
Please see our free downloading policy.



Why so long?

A year or two ago, hardly a day passed without the announcement of some new interactive cable television venture. The announcements promised the consumer a cornucopia: 500 channels of movies, game shows, home shopping and banking, information, and education, all available "on demand," many of them interactive "new media" experiences that would put the customer in charge, no longer a couch-potato "viewer," but a full participant.

Many of these announcements equated interactive television with "the information superhighway," projected vast changes in our daily lives, and spoke of grand consolidations as the giants of the telecommunications industry struggled to find the expertise and the financing to meet the challenge - and to avoid getting left out of the fracas.

In 1993 alone:

Everybody, it seemed, signed joint venture agreements, letters of intent, and sometimes just press releases, with everybody else in the cable, telephone, and computer industries.

Across Europe, national television systems faced falling subsidies, privatization, competition from U.S. cable and satellite systems, and challenges to their legal status (as Deutsche Telekom was challenged by three corporate giants, RWE Energie, the Deutsche Bank, and the Mannemann engineering group). They reacted nervously to the American turmoil, seeing both the danger of yet another American telecommunications invasion, and an opportunity for new revenues.

So where did all the fuss and feathers go? Was it just so much burbling and vaporware? As of January 1995, industry observers counted one bare-bones interactive system operating on conventional hardware (GTE's Main Street), and a mere seven houses connected to a full broadband interactive cable network - Time Warner Cable's experimental Full Service Network in Orlando, Florida, finally up after a year's delay. And those were connected not through market-price set-top boxes, but through custom-made Silicon Graphics computers costing several thousand dollars each.

"There's a difference between being in the announcement business and being in the business," grumps Rich Frank, chairman of Walt Disney Television and Telecommunication. "I've been hearing about this Orlando thing for three years."

Why the delay? Is this really going to happen? If and when it does, will it be worth the wait? What's going on behind the scenes?

We asked a number of people who might know: industry executives, Wall Street analysts, Internet pundits, and eminent professors. Their replies were cacaphonous and often contradictory, an image of the chaos of the industry, but out of that chaos a theme emerged: complexity.

No one has ever done this before. The announcements were somewhat akin to Columbus' presentations to Ferdinand and Isabella about the possibilities of a new route to the Indies. No one knew what the obstacles were until they tried. Now, after a year or more of trying, the obstacles are clearer, larger, and far more numerous than they had seemed at the start. There are regulatory hurdles, technical bottlenecks, and a bewildering array of unsolved financial, competitive, and marketing questions, all of which must be solved before your TV explodes with choice.

Where are they now?

Still, things are stirring. Time Warner has turned the switch in Orlando, and expects to be bringing full services to 4000 homes by the end of the year. USWest has a similar test in Omaha. Southern New England Telephone started a trial with 370 homes in West Hartford, Connecticut, last April, and is now expanding that to 150,000. GTE turned the key on its limited commercial service - which offers such services as home shopping, play-along games, and information, but no movies - back in January of 1993. It now offers the service to some half million homes in New England, but counts only 4000 actual customers. Pacific Telephone (an RBOC) hopes to be able to deliver "video dialtone" capacities to some 1.3 million homes in California by the end of 1996. BellSouth (another RBOC) begins an 18-month field trial of 12,000 homes in the Atlanta area this summer. AT&T and Cablevision (a major American cable company) have announced an intention to bring "video on demand" (VOD) to all of Cablevision's customers this year. And TCI expects to have full broadband interactivity to all customers by the end of next year.

British Telecom is already operating a trial system. Deutsche Telekom and some other European national television systems (including France, Italy, Sweden, and Norway) are likely to start up trial systems within this year.

Singapore Telecom, as part of its push to build a state-of-the-art broadband communications infrastructure, has just announced its own "video on demand" (VOD) trial project, to start up this summer. Singapore is a natural ground for the new technology: All of its telephones are already digital, and all its exchanges are already linked by optical fiber. Ninety-nine percent of the population own color TVs, 84 percent own VCRs, and the entire population will have optical fiber to the home by 2005. Japanese companies Mitsui and Fujitsu are the project's leaders. Yet on their home turf, though they show great interest, Japanese companies so far have signed on no dotted lines to bring interactive television to Japan.

Interactive TV is part of the renascence of the power of the American economy, as American companies such as Hewlett Packard, Sybase, Scientific Atlanta, General Instruments, AT&T, and Oracle lead the charge on supplying the world with the technology to make such systems work.

In fact, many industry executives, and some outside observers, claim that nothing is going wrong, that everything is on track except the public relations war. Phil Anderson, a professor at the Tuck School of Business at Dartmouth, says, "This is not a story about technical failure. It's a story about a failure to manage expectations. This clearly will be revolutionary. It will restructure a lot of economic relations. Investors should be excited. But there is and should be a ten-year lag between what the investment community considers exciting and what shows up in the home"

Glenn Britt, head of Time Warner Cable Ventures, says, "The only mistake we made" in setting up the Orlando trials was that "we were naive in setting the starting date. In fact, we didn't have a particularly strong handle on when it was going to start. It wasn't that we were late in starting, it was that our initial guess was wrong."

Besides, says Disney's Rich Frank, "The first one's not always the right one. Perhaps they're building an Edsel."

Can it be done?

The technical complexities of this new landscape at times seem limitless. Tom Mandel of SRI International (a consulting firm in Menlo Park, California), expresses it as a "plug and play" problem of a gargantuan scale: "Ordering up a network, some ATM switches, set top boxes, video servers, and the like doesn't mean that you can just plug them together and they'll work."

The hurdles between a conventional cable-TV system and a truly interactive switched broadband network are many [see "How It Works"]. They include:

Each of these technologies is advancing rapidly, with capabilities expanding while the price is falling. By late 1994, for example, wholesale prices of hard disks had fallen to about forty American cents per megabyte, while memory chips had come down to $30 per megabyte. At some point, in some configuration, it becomes possible to cram enough power into a set-top box and enough capacity into a system to make true interactive television work at a price that people will pay.

Those complexities, the kind of solutions people come up with, and what assumptions they make about what will be available in a few months or years, directly affect what a system can deliver. Gerald M. Levin, the chairman of Time Warner, has declared, "I want to deliver every movie ever made," yet the limitations of bandwidth, memory, and compression schemes may limit interactive video-on-demand systems to the current hits, rather than making available vast libraries of cinema.

And they could play havoc with business models: it may be possible to deliver services that people like, but not at a cost that people like. Pacific Bell, for instance, is testing whether consumers have a "bit-rate threshhold," a minimum level of quality they will accept in a digital signal for which they are paying extra. The higher the bit-rate, the more information coming down the cable every second, the higher the quality of the movie on the screen. Six megabits per second of digital video gives a picture quality equivalent to an analog cable signal. "Three megabits, where we are working now," says Kevin Seeman, director of broadband services for Pacific Telesis Video Services, "is better than what we see out at the end of our analog cable runs. If we can get it down to a 1.5 megabit rate, it would have a very large impact" on the viability of the project. If people insist on six-megabit quality, "we could find that [the concept] doesn't work" as a real business.

Much of this enormous ferment has to do with Moore's Law, the widely-quoted axiom that information technology generally doubles its power and halves its price every 18 months. Interactive television networks must meld together compression technology, massive computing power, vast amounts of computer memory of many types, thousands of miles of fiberoptic cable, and highly advanced video server technology. All of these must be highly efficient, as error-free as possible, and cheap. Today, few if any of these technologies can deliver the goods at a price that makes commercial sense for the consumer. But, according to Greg Hoberg, manager of video solutions for Hewlett Packard, "We expect large drops in prices in all these elements during 1995 and 1996."

That's the bet: will the technology improve fast enough, and the price drop fast enough, to make these systems commercial? Since it takes years to build these systems, a company that waits until the technology is ready will be buried by the competition. But a company that jumps the gun could lose tens of millions of dollars.

In Orlando, Time Warner has shaped a powerful strategy around Moore's Law: they are doing whatever they need to do with today's technology, using what are essentially customized Silicon Graphics Indigo computers - powerful and expensive graphics workstations - as set-top boxes. So they can offer a full array of services with today's technology, though at a price that is costing the corporation tens of millions of dollars. Silicon Graphics hopes to have all that power reduced to a single chip by 1997. Time Warner is betting that, by the time that happens and they can produce a set-top box at a reasonable price, they will have gained incalculable experience and information from the trials, and be ready to roll out interactive networks nationwide. "It's the right strategy," says Anderson. "You have to bet on what you think the world is going to be like two or three years in the future."

How do you make money at this?

How do you build a business plan for a business no one has tried before? Donna Hoffman, a professor at Vanderbilt University in Nashville, Tennessee, a specialist in marketing for the new media, states flatly, "There are no data showing that you can make money at this yet."

"Until the results of the tests come in," says Mandel, "no one really knows what consumers will do with the new services."

Most of the business models for these ventures take a "build it and they will come" attitude, assuming that consumers will flock to shop, play games, and especially download movies over interactive television. But according to Mandel, the amount that consumers spend on information of all types has grown only slowly as a percentage of gross domestic product for over half a century, so if these new media are to win any market share, they may well have to wrestle it away from the media already in place, from newspapers and broadcast television to CDs. However, he cautions, "Worldwide and in Europe, the market is going to grow a lot, and experience gained in the U.S. market will give American companies a competitive advantage in the huge global market that will emerge over the next 10 to 50 years. The Americans, the Western Europeans, and the Japanese, plus a few others, are the 'early adopters' of the global media consumer society. But the real growth in the long run is going to be elsewhere."

Previous interactive TV experiments (such as the QUBE trial in Columbus, Ohio) did not show much promise, and pay-per-view events over traditional cable have rarely had takers of more than single-digit percentages of the audience. Preliminary market surveys deliver conflicting evidence. Many business plans assume that downloadable movies ("video on demand") is the "killer ap," the application that will drive the consumer to adopt the system, followed closely by other forms of entertainment (such as games) and home shopping. But a survey of 600 Americans by Macworld, a computer magazine, showed movies ranked only tenth out of 26 possible uses of the "information highway," with other shopping and entertainment uses clustered near the bottom of the list. People were much more interested in going to school, voting, or using reference materials online. Three quarters of the people surveyed wouldn't pay more than $10 per month for any service except for at-home education.

But a much larger Associated Press survey commissioned by the three major American broadcast networks and the National Association of Broadcasters showed an opposite result: "Of proposed interactive services, 30 percent of respondents were interested in 'movies on demand.' No other service offer came close: Information got a 12 percent response, sports got 12 percent, video games 11 percent and home shopping 7 percent."

In June 1993, just as the interactive television ballyhoo machine was getting into high gear, SRI International released a study of the industry's prospects, backed by $1.5 million in research sponsored by 50 big "new media" companies. The verdict was harsh: except for games, "no residential application will achieve greater than 10 percent penetration of U.S. households through 1997." Downloaded movies were a particularly dubious bet: "less than 2 million U.S. households will be subscribing to VOD services in 1997." SRI considered its view "realistic." Almost everyone else called it "pessimistic." The Wall Street Journal called it a "freak show." But according to SRI's Steve Krause, since the report was issued, events "have brought the mainstream very close to our view," that "this stuff is technically harder than most people originally thought," and that the demand for it was far from certain. "As far as we can tell," says Krause, "we're no longer a freak show."

The video rental and sales industry, which takes in $18 billion per year in the United States, does not seem convinced that interactive television will hurt them, at least not soon. Two years ago, Wayne Huzinga, chairman of Blockbuster Entertainment, the largest American chain of video stores, told Business Week: "[Interactive television] gives the appearance of happening very quickly, but it will be a long time before the average family has it." Tom Grieb, general manager of GTE's Main Street, agrees: "To really embrace this, people have to change their behavior. That usually takes a long time. It's just like ATMs [automatic teller machines], which were widely held to be a failure in mid-70s."

And his potential competitors, to some extent, agree with him: "We don't believe we're going to put anyone out of business," says Seeman of Pacific Telesis. "We do believe we can bring the consumer some convenience - we're never out of stock, and you don't have to return the tape."

Grieb, whose system does not carry VOD, doesn't think interactive systems need it: "We don't believe there is a 'killer ap.'"

Home shopping, the second candidate for "killer ap," is also coming into question in the United States. Fingerhut has canceled its "S Network," and Federated Department Stores has pulled the plug on "TV Macy's," declaring that the technology and cable access were not ready for it. Saks Fifth Avenue gave up on their experiment selling clothes over the QVC cable shopping channel after only a few shows, because it cost too much, and too many people returned the merchandise. Even Gerald Hogan, CEO of the cable-based Home Shopping Network, put general use of interactive electronic retailing a decade in the future. At a retailing conference in New York, Time Warner's Gordon Cooke was forced to declare: "History tells us that ATMs, fax machines and cellular phones were all initially considered failures. Electronic retailing will become a force to be reckoned with."Retail analyst Jeffrey Feiner of Saloman Brothers says, "The American consumer still likes to shop, to go to the mall and touch and feel the merchandise."

Finally, there are those who believe that we will never know whether people want or will pay for interactive television until we can tell them what's on: the story's the thing that will drive all marketing trends, not the technology. As Disney's Rich Frank put it, "All these geniuses can figure out how to deliver anything they want. But if all they can find to deliver is 25 versions of the Amy Fisher Story, no one's going to watch it."

But the experiments will go forward, and the networks will be built, at least in the United States, whether or not there is a market for interactive cable television - because most of the investment is not about interactive cable television, it's about the real cash cow, local telephone calls.

American telephone companies are replacing their "twisted pair" copper-wire networks with broadband coaxial cable, glass fiber, or a combination of both, so that they can improve their telephone service and offer a broader range of data services into the next century. Providing "video dialtone" is just another service that they could offer on a network they are going to build anyway. It's what John Malone, chairman of TCI, characterizes as "a lucky strike extra."

American cable television companies, on the other hand, really want to get into the local telephone business - POTS, or "plain old telephone service" - whose cash has been the engine of the enormous growth in regional telephone companies over the past decade. They already have the broadband cables in place, but to become a telephone system they have to change their architecture from a "tree and branch" system, with everything coming from one central office to individual sets, to a "switched network" system, in which everyone, potentially, is connected to everyone. "We would certainly be in a position to provide telephony," Malone told Wired last year, "at substantially lower prices than you're currently being charged." As proof, he pointed to TCI's success in competing against British Telecom in the United Kingdom. Once cable companies have changed the architecture to a switched system, providing interactive cable television becomes easy, a "value added" service that they will have to do so that they won't lose customers to their new competitors, the telephone companies.

What are the rules?

Against this background, the American regulatory and contract battles become clearer. SRI's Mandel calls this problem, "Who can do what, and what will the government require you to do?"

The problem is most acute in the United States, where telephone networks, cable systems, television networks and stations, and the studios that made movies and television programs, have all been almost completely private, separate, and heavily regulated under vastly different asssumptions and bodies of law. A joint venture between, say, a movie studio and a local telephone company to provide video over the same lines that the telco uses to bring its customers telephone service would involve creative compromises not only by the Federal Communications Commission (FCC), but also by the public utilities commission of each state in which the venture operates.

When cable and companies both morph into switched, broadband networks, each eager to get into the other's business, "the battles will be long and bitter," says Dartmouth's Professor Anderson. Each industry will use the regulatory process and the courts to pierce the walls around the other industry, at the same time that it defends its own turf.

"We're still waiting for our '214,'" says Seeman at Pacific Telesis, referring to the "construction permits" the industry must get from the FCC - usually a routine matter, unless you are talking about video dialtone. Of those requests for permits, as Seeman puts it, "many have gone in [to the FCC], few have come out."

Even friendly companies can be tied in knots by the regulatory uncertainty. Disney announced early last August that it would form a joint venture with three regional telephone companies (Ameritech, BellSouth, and Southwestern Bell) to develop interactive video services. They're still working on the contract. "It's a 300-page document as dense as anything I've ever read," says Disney's Rich Frank. "We are trying to write agreements between companies in the hardware business and the software business about a business that isn't there yet, while the government is changing the rules. So we have to keep saying, 'If this happens, we'll do this, if that happens, we'll do that.' It's one thing to get into a business that you know is there. This business comes with giant trademark issues, copyright issues, and [writers, directors, and actors] guild issues. We're trying to write agreements about navigators have not been designed yet, and work out marketing plans for products that don't exist. We know we won't get it right, but we have to do our best."

But, says Anderson, at least the regulatory part of that complexity may suddenly disappear. The key sign is that many of the telephone companies are going outside of their own territory to become video carriers elsewhere, and they end up on the other side of the same fight. "If you're a phone company, like USWest, you want to be as ornery as possible with state rate commissions, to keep other companies out. But at the same time, operating as a cable company, they have bought much of cable market in the eastern seaboard states. They are having an increasing amount of trouble arguing different things in different states. For the 'Baby Bells,' the contradiction between defending their territory, and cooperating with cable companies outside of their region, cannot persist." Indeed, in some areas, such as Rochester, New York, the local phone company has dropped its objections to letting the cable company into its telephone markets, in return for being allowed to branch out into broadband video services.

Is it even the right model?

All roads lead to a convergence of the television, the computer, and the telephone into one device hooked into a vast network, pulling down information, games, entertainment, and contact with other humans. But there is little agreement what that network will look like. Put simply: will it be a fancy version of TV? Or will it be the Internet made simple, with video, sound, and commercial programming thrown in?

Or, in info-jargon, will it be "one to many," like television broadcast from a central source to many receivers, or will it be "many to many," like the Internet, a network in which every node can be a producer as well as a consumer.

The designers and builders of these new cable systems "don't know what they are doing," says Professor Hoffman at Vanderbilt. "As long as you stick with a one-to-many model, you're missing the boat. What the people who are running these systems need to know is that people don't come to them for choices, they come for deeper psychological reasons." She cites studies showing that the reason people watch shopping networks is not for shopping convenience but for affilitation - they identify with the hosts and the other members of the audience. For instance, the popular computer network America Online has huge retail sections, "but few people use them, and the system earns a minuscule amount of its income from the shopping venues. People spend the majority of their online time on email, chat rooms, and forums. They want to talk to each other. People are searching for a sense of community. An important feature of computer interactions such as the Web and the Internet is that the user can be not only a consumer of information, but also a provider. Current implementations of so-called interactive multimedia like interactive TV do not offer that." Allowing people to choose their package of pre-produced entertainment "is not interactive TV," says Hoffman. "It's just new and improved TV."

"The set-top box, and what it will carry, is only the tip of the iceberg," says Anderson, from Dartmouth. "The real story is the interactive multi-media network that converges near the TV. It will be a mixture of the different models that we have now. There is no question that the many-to-many model will ultimately prevail. The question is only how long that will take. By 1997 or so we will have legitimate voice recognition. I expect that, eventually, many-to-many videoconferencing will be the most popular use of these networks. The answer is not very far down the line. As soon as it is offered it will become more popular than home shopping. Anyone who overlooks the power of the thirst for community has a very neanderthal way of looking at the network."

The numbers offer some support for this point of view. While interactive cable stumbles through trials trying to find hard numbers that justify the enthusiasm of its investors, the Internet continues to grow in double digit percentages every month. The World Wide Web, largely unknown only one year ago, has exploded as easy-to-use interfaces such as Netscape and Mosaic have become commonplace. Between 1989 and 1993, the percentage of U.S. computers tied to some kind of network climbed from 10 percent to over 60 percent. While the sales of video game systems declined slightly in 1994, with customers grousing about the $300 to $500 price tags on the newer versions, the sales of multimedia-equipped personal computers doubled to over 7 million units in the U.S. - at four to five times the price of the game machines. For the first time, Americans shelled out more dollars for personal computers than for color television sets. In 1993 and 1994 alone, people bought an estimated 100 million new personal computers worldwide. Industry experts expect that within the next 12 to 24 months widespread use of cable modems will allow computers in the home direct high-speed access to the Net for a nominal fee. Most interactive cable services will cost the consumer more than today's cable television. But by Moore's law, every seven years, most computer functions grow 100 times in power while dropping to one hundredth of their former cost. These two trends are bound to cross.

It may be that by the time fully interactive broadband services arrive in the average home via cable networks, they will already be there via computer networks.

(Donna Hoffman may be reached through her web page at Vanderbilt University.)

Technology | Articles | Main Page