Forget the browser. (Bill Gates has.) Microsoft wants to be in the box.
Wired 6.04 ⎢ April 1998
BILL GATES AND FOOTBALL MAY NOT EXACTLY BE LINKED in the public eye, but judging from his performance at January's Consumer Electronics Show, this is something Microsoft's CEO would like to correct. It was there that he announced his landmark deal with Tele-Communications Inc. — an agreement that will put Windows CE, the lite operating system, on 5 million digital set-top boxes which the cable giant plans to deploy sometime in 1999. Demonstrating the television of the future in the gizmocentric air of the Las Vegas Convention Center, Gates — the richest nerd in America, and certainly the most vocal in his disdain for TV — showed how you could watch a football game, summon statistics, and even order yourself a jersey at the same time.
Gates, of course, knows that for millions of people, television and football are indelibly linked — and if he has his way, television and Windows will be equally inseparable. Even as he spoke, Microsoft was finalizing a deal with Gemstar, maker of VCR Plus, to license its interactive program guide for the forthcoming Windows 98. But Gates's real focus is on the set-top box, the noisome little gadget that serves as a halfway house for the signals coming into your television set, and he's not alone in his interest. Less than a week earlier, Sony had announced a deal that would put its brand — one of the best known in the world — on many of the CE-equipped boxes General Instrument Corporation is building for TCI. The previous day, Sun Microsystems, Microsoft's archrival, had announced an agreement to supply TCI with a scaled-down version of Java for those same boxes — an arrangement that was immediately perceived as a blow to Microsoft. So eager was Microsoft to have something Gates could announce at CES that its negotiating team, which had been in discussions with TCI since last April, had worked well into the night to cut the deal. Not since his long-ago talks with IBM — the ones that made MS-DOS the standard operating system for the PC and eventually allowed Microsoft to dominate personal computing — had Gates been so dependent on a potential partner.
"Microsoft desperately wants to be in the box," says Gary Arlen of Arlen Communications, a research firm based in Bethesda, Maryland. "This was TCI's way of saying, We still control the real estate. From Microsoft's point of view, at least they've gotten in. But they don't control the box — yet."
If cable has its way, Microsoft never will. "The industry is very wary about going down an approach like the PC industry's — a proprietary operating system coupled with a CPU," says Bruce Ravenel, TCI's Internet services chief, who was instrumental in the talks with both Microsoft and Sun. "Because that places control in the hands of the suppliers of those components." But unlike Netscape and its allies in the computer industry, TCI has no interest in bashing Microsoft. Far from it: even without the billion-dollar investment he'd been expected to receive, TCI chair John Malone gave Microsoft a hearty endorsement at CES, saying "This is the horse we're going to ride" and declaring it furthest along on the path toward PC-TV convergence.
What really came through in Vegas, however, was the size of the arena Microsoft has entered. This is no grudge match with puny Netscape; more like a tango among titans. With US$11.4 billion in sales last year (up from $4.6 billion in 1994), Microsoft is facing companies like Sony ($46 billion), Time Warner ($21 billion), and TCI ($8 billion, not counting its joint ventures with half the media conglomerates on the planet). At stake is not just the future of the American living room, but global lifestyles in the 21st century. As Microsoft's war with the Justice Department over the marriage of its Internet browser with Windows 95 devolved into a staredown, with high-powered attorneys trading epithets like 6-year-olds, suddenly it was possible to think, Browser? What browser? Was the epic Netscape-Microsoft-Justice Department confrontation just a sideshow, a loss-leading, brief-spewing, lawyer-saturated warm-up to the main event? Might the browser wars be a footnote by the end of the century?
Check the arithmetic: PCs are holding steady in fewer than 40 percent of American homes. Only 20 million homes have PCs with modems. Some 68 million have cable television. For Bill Gates, cyberspace isn't where the action is. To maintain its phenomenal growth, Microsoft needs to be in the market its people call "televisionspace." The question is, How big a chunk can it claim?
"MICROSOFT HAS BEEN IN TELEVISIONSPACE FOR BETTER THAN FIVE YEARS," recalls Mike Conte, group manager for digital television, sitting at his desk in Building 9 of the company's Redmond headquarters. "We've had a lot of history. Some good experiences and some" — he pauses momentarily — "some learning experiences, shall we say. But now we've come full circle."
The circle began with a memo written in the summer of 1991 by Nathan Myhrvold, Microsoft's chief technology officer. Myhrvold wanted to take the computer beyond the PC — particularly into interactive television, which entranced him with its possibilities of video-on-demand and, by the way, instant obsolescence for Blockbuster and the VCR. An Advanced Consumer Technology Group was formed, almost as an in-house start-up, to develop software that would control a variety of non-PC devices, digital TV sets among them.
To head the effort, Gates and Myhrvold recruited Craig Mundie, founder and CEO of Alliant Computer Systems, a Boston-area company that had just gone belly-up trying to build massively parallel supercomputers — machines relying on thousands of microprocessors — for the scientific-research market. Mundie, a onetime Data General executive who'd helped found Alliant in 1982, was an interesting hire: smart, focused, intense, the kind of guy people regularly describe as "visionary," well versed in operating systems but with no background in Windows. Supercomputers were a long way from what Microsoft had in mind, but Mundie thought his team-building experience at Alliant could be put to good use. He quickly junked Modular Windows, a stripped-down operating system that had been at the heart of an abortive interactive CD player Microsoft had recently introduced with Radio Shack, and decided to start fresh. The result was Windows CE — a name that presumably refers to "consumer electronics," though Mundie insists it has no meaning whatsoever.
By 1993 Microsoft was in talks with TCI, Time Warner, AT&T, and other telecom giants about a joint venture for interactive television. The talks went nowhere because, as Randall Stross reported in his book The Microsoft Way, both TCI and Time Warner insisted on a 50/50 ownership scheme that would tie Microsoft to the project but leave them free to walk. Nonetheless, Microsoft and TCI did join forces on a series of interactive television trials that were to run in Seattle in 1995. The trials never got beyond the in-house stage, but Windows CE evolved into a product that could drive everything from digital telephones to miniaturized computers. By late 1996, when version 1.0 finally shipped with the release of the Handheld PC, several hundred developers had put it through any number of rewrites at a cost of hundreds of millions of dollars. "It takes a village to raise a child?" says Conte. "It takes a couple of villages to make an operating system."
But Mundie and his team didn't just have to develop an operating system; they had to reinvent the computer. From the dawn of the PC era — which is to say, from the inception of Microsoft — the computer has been a beige box with a keyboard and a screen and, in recent years, a mouse. What if it weren't beige at all? What if it were sleek and gray and sat on top of a TV set? Or if, as Gates wrote in his 1995 blueprint The Road Ahead, it were no bigger than a wallet, equipped to send email and faxes, and charged with digital money? Would Microsoft still supply the software?
Mundie and his group had to rethink the company as well. As the railroads once had to decide if they were in the train business or the transportation business, Microsoft had to decide if it was in the personal-computer industry or something bigger. With Myhrvold and Gates around, that was a no-brainer. The real issue was how the company could profit as the computer morphs into myriad new forms and insinuates itself into people's lives in ways they barely perceive. The economics of the PC business were by now well established, and on Microsoft's terms. But as digitization hit the telephone and the television, Microsoft would increasingly find itself on turf owned by bigger companies — AT&T, Time Warner, TCI. What could Redmond bring to the party, and what could it collect in return?
As they grappled with these issues, Microsoft's people also realized they'd have to reinvent TV — not an easy task from the cloistered, sylvan setting of their Redmond campus. A brainy corporation reliant on faceless, voiceless communication through electronic blips, Microsoft would have to transform a showbizzy medium best known for inducing mind-numbing passivity into one that somehow encouraged the audience to talk back. How would it do that?
The Big Picture
Microsoft moves beyond the desktop.
★ owned by Microsoft
★ Aspect Software Engineering
Mundie and his group tried to provide the answer with "My TV" — Microsoft Interactive Television, or MITV for short. But aside from the catchy acronym and a video server known as Tiger (because it sliced data into "stripes" for storage), MITV didn't have much going for it. Demonstrating Tiger for an audience of utilities executives at the Edison Electric Institute in June 1994, Gates got a disk-error message as his screen went blank. "I think that's the end of the demo," he quipped. "You can see it's not going to happen overnight." A splashy announcement at a cable-industry trade show a year later, with partners ranging from ESPN to Starwave to the Home Shopping Network to Samsung Electronics, failed to impress the pundits, who noted that Microsoft had mainly succeeded in porting vaporware to the television environment. As for the cost, the best anybody could say was that it would be affordable someday.
Of course, nobody else knew how to make interactive television work, either. Time Warner was demonstrating that at its trials in Orlando, Florida, where 4,000 cable subscribers were getting online games, all available channels plus movies on demand, and pizza by remote control, courtesy of high-powered video servers from Silicon Graphics and set-top boxes that cost $8,000 apiece to produce. The technical difficulties were awesome: devising computer architectures, managing data flows, keeping costs out of the stratosphere, writing software to handle it all. The pizza-ordering application alone took programmers months of work and hundreds of thousands of dollars to create.
But while Microsoft was focused on the distant promise of digital television, the Internet — the ready-made connection with its anticommercial, up-from-under flavor — exploded into the consciousness of Wall Street and Silicon Valley and the media. By mid-1995, with the Internet boom in full motion, interactive TV was fast becoming a joke. To protect its core business and retain its credibility, Microsoft had to do something fast. So even as he was touting MITV to the cable industry that May, Gates was circulating an internal memo called "The Internet Tidal Wave" in which he announced that the Net was the most important development in computing since the début of the IBM PC and declared it the company's Number One priority.
Microsoft was still committed to three sets of interactive television trials — in Seattle with TCI, in suburban Dallas with SBC Communications, and in suburban Tokyo with Nippon Telegraph and Telephone. Mundie was predicting rollouts in major metropolitan areas during 1997. But by the end of 1995, all the trials except Tokyo had been put on hold.
That December, Gates held an "Internet Strategy" briefing for 300 analysts and reporters at Seattle Center, in the shadow of the retro-futuristic Space Needle, Seattle's answer to the Eiffel Tower. Despite its phenomenally successful release of Windows 95, Microsoft was on the defensive because Netscape had beaten it to the Internet. Its much-hyped Microsoft Network was being retrenched as a node on the Net, just as Netscape cofounder Jim Clarke had predicted; much worse, its hold on personal computing was seen to be in doubt. So Gates told his audience that Microsoft was "hardcore about the Internet," that it would "embrace and expand" the Internet. Amid the hoopla, few took note when he also asserted that interactive television remained "the holy grail."
ASIDE FROM THE WORLD'S MOST expensive pizza-ordering system, what did interactive television yield? "That it takes a hell of a lot of money and time to change national infrastructures," says Mundie. Also, some experience in price points: the $8,000 set-top was not an option. But the Internet has proved to be a far more efficient schoolhouse. "The initial interactive television effort didn't have the answer to two key questions," says Hank Vigil, Mundie's strategic planning chief, who was brought in from Microsoft's electronic-commerce group in the fall of 1995 to salvage what he could. "What were the business models that would excite consumers enough to get them to make good on our investment? And where were these services and applications going to come from?" The Internet had services and applications galore, and plenty of people ready to use them — which meant, says Vigil, "You could break the chicken-and-egg problem."
From Microsoft's perspective, however, the best thing about the Internet may be that it spawned WebTV, the Silicon Valley start-up that brought the Internet into America's living room. It started when a onetime Apple exec named Steve Perlman happened upon the Campbell Soup Web site one night and decided to see if he could make it come in on TV, where people who like Campbell's soup might actually see it. His scheme worked, so he formed a company that by the fall of 1996 was generating major buzz. Shortly before WebTV was ready to launch, David Cole, a VP in Mundie's Consumer Platforms Division, paid Perlman a visit — with the result that Microsoft, Bill Gates, and cofounder Paul Allen all became backers. Later Mundie himself met with Perlman and had what he describes as "an epiphany": the realization that he and Perlman had virtually identical visions of how digital television would emerge. Mundie spoke to Gates and to Paul Maritz, group vice president for platforms and applications; within weeks, Microsoft bought WebTV in a $425 million deal.
For Microsoft, WebTV is the ultimate interactive television lab: not only does it sell a $199 set-top box that displays Web sites on ordinary TV sets, it has more than a quarter-million subscribers who spend (according to company research) two and a half times as much time online as PC users. And this is no early-adopter crowd: 60 percent of them don't even own a PC. "The funny thing is, it really is a computer," Conte says of the WebTV box. "But people think of it as being better TV. That's a model for us to learn from."
While Vigil was peeling off pieces of MITV — like NetShow, the interactive video program — and directing them toward the Net, Mundie was laying the groundwork for phase two of the assault on television. The timing was critical: advances in microelectronics were finally bringing video-on-demand to the realm of affordability, and digital television was happening at last. Mundie, outgoing and upbeat, was the perfect guy to articulate the Microsoft vision: "He can excite people about it," says Vigil. "In particular, he can excite senior executives at other companies about it." He got his chance in April 1997, when the CEOs of the major US cable companies went on a "spring technology tour" that took them through Redmond and Silicon Valley. Three months later in New York, speaking to the same CEOs in the boardroom at Time Warner headquarters in Rockefeller Center, Gates filled in the details of how Microsoft could provide their one-stop technology solution.
If there was a precise moment when Microsoft lost the chance to get the lock on digital television that it has on personal computing, this was it. Weeks earlier, Microsoft had bought a $1 billion stake in Comcast in the apparent expectation that this would give it an edge in the software sweepstakes. Now Gates was proposing a system with Windows NT on the servers and Windows CE on the set-top boxes, and Microsoft collecting not just a licensing fee, but a cut of what the cable companies would charge for everything the boxes did. His sales pitch became a wakeup call. "It didn't go over too well," recalls TCI's Ravenel. "The cable industry never finds it easy to sell its future — especially when it's not sure what the future holds."
A MONTH AFTER GATES'S PRESENTATION, MALONE stood up at TCI's annual meeting outside Denver and said, "Bill Gates would like to be the only technology supplier for this whole evolution. We would all be very foolish to allow this to happen." Instead, Malone and his cohorts set up Open Cable, an initiative to ensure that digital set-top boxes would be compatible with one another whatever their manufacturer or operating system. Administered by CableLabs, the industry's Colorado-based research-and-development organization, Open Cable was intended to make it impossible for any software company to dominate digital television. Malone made it clear he wanted a "layered" approach, with a variety of companies providing different portions of the system. This, he told reporters after the shareholders meeting, is something Microsoft "will just have to accept."
It certainly fits with the way cable companies have always operated. "If you follow the set-top market, there are two major players," says an executive at Scientific-Atlanta, the Number Two manufacturer behind General Instrument. "The cable industry likes having them compete with each other." As it happens, Scientific-Atlanta has an operating system of its own: PowerTV, developed by a Silicon Valley group that went looking for venture funding and ended up a subsidiary. Recently, Scientific-Atlanta received an order from Comcast, the fourth-largest cable operator in the US, to supply its new Explorer 2000 digital set-top boxes running PowerTV to Comcast subscribers in Baltimore — and never mind about Microsoft's investment in Comcast. Gates's vote of confidence gave a huge lift to cable stocks, Comcast's among them, but he doesn't seem to have reaped any other reward for his largesse.
While Scientific-Atlanta was backing PowerTV, General Instrument was in talks with both Microsoft and Sony. The deal it announced in January, which calls for Sony to lay out $188 million for 5 percent of the company, is only preliminary; the real talks are about developing the next generation of set-top boxes, which Sony executives think could form the basis for a network architecture that would manage all the datastreams that go into the home — radio and television broadcasts, cable and satellite transmissions, the Internet, even packaged media like DVDs and compact discs. Sony's interest in this kind of digital integration is one reason it bought into Japan Sky Broadcasting, the satellite consortium organized by Rupert Murdoch's News Corporation. In the US, where cable is dominant, it makes sense to focus on cable. But Sony's architecture wouldn't have to be centered on the set-top box; it could be inside the TV set, in a PC, or in the basement where nobody ever has to look at it. Nor does it necessarily pit Sony against Microsoft, since Windows CE or any other operating system could lie on top of the network architecture.
But the real problem for Microsoft is that operating systems for digital TV just won't be the big deal they are in PCs. When Malone said Microsoft was furthest along on the path toward convergence, it wasn't Windows CE he was talking about; it was WebTV, which will provide the graphics processor and other hardware for TCI's new set-top boxes. "It doesn't matter what the CPU or the operating system is," says Ravenel. "That's just the guts inside the box that makes it work right. The services are what matters."
Microsoft knows this only too well; that's why it's still trying to sell the cable industry on offerings like Sidewalk, its interactive city guides, and Expedia, its online travel service. (And the Internet? "We view the Internet as one of the 'features' of digital TV services in the future," says Mundie.) Microsoft doesn't stand to make much money on the TCI deal: the licensing fees are said to be under $60 million. It's hoping to use Windows CE and the WebTV technology as leverage to pry open the box so it can get at the revenue streams inside. Yet its tactics, in cable as with the Justice Department, suggest it has a better understanding of its own wants and needs than of the world it's charging into.
TCI and the other cable operators see themselves as buyers and resellers of video content; they're not about to share their subscriber revenues with a software outfit. "It's a whole new paradigm — a network paradigm," says a senior executive at Comcast. "The network operator provides consumers with the operating system and with the applications that we've contracted for." "Cable companies don't care about operating systems," concurs the Scientific-Atlanta executive. "They care about services they can charge the customer for. Being a bit pipe is not the way they view their business."
On the other hand, being a bit pipe is exactly the way the phone companies view their business, which is probably one reason why Microsoft was herding them onto the information superhighway even as it cut its deal with TCI. The cable guys don't seem worried: "Tele-TV and all that stuff — that's history," says Ravenel, referring to the telcos' defunct TV-programming consortium. "We'll compete with them in lots of businesses, including the phone business someday — but we don't see them coming into the video business."
For the cable companies, video — not interactivity — is what it's all about. In the early '90s, the exorbitant cost of digital set-tops and servers seemed to require all the gee-whiz options that interactive television promised and failed to deliver. But the trials showed that what consumers really want is more TV — the TV they want, when they want it. TCI's new boxes will cost only $300, less than its current, first-generation digital set-tops, and will be provided to customers for a small monthly fee. "Interactivity is icing on the cake," says Ravenel. "We don't have to make a bet on it — we just make a bet on what we know now, which is our cable business."
But more has changed in five years than the cost of set-tops. Microsoft is stronger than it's ever been, while the cable industry has only been brought back from the dead by the boldness of Gates's bet. Microsoft's stock was worth $212 billion at last report, while the market capitalization of the three companies that make up TCI is only $34 billion — which means that Malone's entire empire is worth $7 billion less than Bill Gates personally. Executives at TCI, whose 14 million subscribers make it the largest cable operator in the country, say they can finance the necessary upgrades internally, but with $15 billion in debt, that's not going to be easy. Meanwhile, Microsoft is sitting on a cash hoard of $9 billion-plus. But don't expect it to relax.
From Redmond, the world looks as threatening as ever — threatening, and challenging. "We're suddenly a little, tiny fish in a much bigger ocean," says Conte. "It's humbling, but also very exhilarating. You know, somebody once said that the beauty of geometric growth is that everything that's happened until now doesn't matter. I think that's true in televisionspace, too. And that's very exciting." ■
Frank Rose is the author of The Agency: William Morris and the Hidden History of Show Business. He wrote "Sex Sells" in Wired 5.12.