Saving TiVo IT'S BEEN WRITTEN OFF BY THE EXPERTS. BUT THANKS TO A SILICON VALLEY SOFTWARE LEGEND, THE COMPANY THAT INVENTED DIGITAL VIDEO RECORDING SAYS IT'S ABOUT TO CHANGE THE WAY WE WATCH TV--AGAIN.
By Kim Girard

(Business 2.0) – Obituaries for the nearly departed TiVo Inc. have been written up for months now, all ready to go when the sad day comes.

They're poignant, these eulogies. There's the part about how feisty little TiVo created a beloved and revolutionary product--the digital video recorder--on a shoestring. The word TiVo, like Google and Xerox and only a handful of other product names, went on to be used to describe what the gadget does--in this case, learning what you like to watch, recording similar stuff for later viewing, and allowing you to pause or otherwise time-shift live action. TiVo promised to transform television, advertising--hell, the culture itself, not least by sparing humanity from having to sit through commercials. Alas, it burned through $567 million between 1999 and mid-2004, and was run down by huge and ruthless competitors that mimicked its technology. But take heart, the eulogists conclude: TiVo will always be with us--as a verb, if not a company.

There's at least one problem with that scenario, however. His name is Arthur van Hoff. He's an obscure but revered high priest of software coding. And he thinks he's devised a way to pull TiVo back from death's door.

In January, TiVo bought van Hoff's tiny startup, Strangeberry, for an undisclosed sum. Hardly anyone noticed. But for TiVo, the deal is shaping up as a masterstroke. Van Hoff is known in elite computing circles as one of Silicon Valley's most brilliant minds. Prickly? A bit. Nerdy? You bet. But, among other career highlights, van Hoff was one of the chief developers of Sun Microsystems's Java programming language, now seen as being among software's most momentous technical feats. Even Sun co-founder and industry legend Bill Joy calls van Hoff "a great programmer." Marimba founder Kim Polese, who worked with van Hoff at both Sun and Marimba, calls him a "giant in the software world."

More to the point for TiVo, van Hoff is an absolute nut for its technology. He owns five TiVo boxes. At Strangeberry, he and a fractious posse of coders put their minds to building the elusive dream machine that TiVo and every other convergence player has long sought: the single, elegant, indispensable device that will control all the elements of home entertainment, from computing to music to movies.

A number of analysts and other third-party observers who've seen the fruits of van Hoff's labors believe that his team may well have succeeded. Strangeberry software gives users the power to do things that no other set-top box or PC has been able to do. You can stream any content from the Net, watch it on your TV, or route it wirelessly to any other device--MP3 player, PDA, laptop. It can all be done with the ease that TiVo's 1.6 million subscribers already have come to relish: You'll never need to click more than a button or two on a single remote to pull entertainment into any room in your house. "Nobody else has technology that comes close," says Daniel Ernst, managing director of New York investment bank Rodman & Renshaw.

And perhaps most significant, TiVo has a fortress of intellectual property protecting the new technology, meaning that rivals will find it more difficult to quickly ape TiVo this time. That could give it enough breathing room to execute a dramatic and difficult shift in its business model. TiVo expects snazzy new features enabled by the Strangeberry software to goose sales of its own boxes. But it wants to pull back from the brutal, margin-sapping hardware game. Instead, it envisions licensing van Hoff's software to some of the very cable companies and other box makers that have been pounding TiVo into the dust, making Strangeberry the software standard for the convergence revolution and guaranteeing it a defensible--and lucrative--position in the battle for the living room.

Is TiVo dreaming? The new approach clearly remains something of a Hail Mary pass. TiVo's rivals, ranging from Microsoft to Time Warner (the parent of this magazine) to satellite-TV operators to consumer electronics makers, all have similar projects in the works and vast resources to develop them. TiVo recently took another body blow when DirecTV, which offers TiVo as part of its satellite service and is a big contributor to the company's revenue, announced that it will also begin offering a rival's DVR technology.

Still, for the first time in many moons, there's a quickening of TiVo's pulse. Its revenue rose almost 50 percent to $141.1 million in the fiscal year that ended in January. Its net loss of $32 million was painful, but less than half of last year's figure. Mike Ramsay, the company's CEO, says subscribers will double to more than 3 million early next year and hit 10 million by 2007. Analysts predict that the company will turn profitable by the end of next year. "TiVo is executing brilliantly," Ernst says. "There's a strong case that this company is going to survive."

It's largely van Hoff's technology that will prove that case, one way or the other. Last year, in a burst of enthusiasm not uncharacteristic for TiVo fans, Federal Communications Commission chairman Michael Powell called its device "God's machine." Van Hoff's work, more than anyone's, will determine whether TiVo meets its maker anytime soon.

"You're calling me lucky?" snaps Van Hoff from across a table in a crowded Silicon Valley cafe. Van Hoff, 6-foot-4 and built like a bull, doesn't like being asked how he, of all the world's coders, wound up a key team member during the early days of Java at Sun. "It's all skill," he says, his gaze steady behind round metal frames. The 41-year-old programmer won't hear anything about serendipity, alignment of stars, or pure luck.

Sun scientist and Java inventor James Gosling heard about van Hoff through colleagues in 1993, while the Dutchman was still earning his master's degree at Scotland's prestigious Strathclyde University. The son of an entrepreneur, van Hoff started coding with punch cards in high school during the 1970s. By college, he already had a reputation for creating from "whole cloth," meaning he'd take a project from idea to architecture to execution, a rare talent for a programmer. He is also fast, known for humbling even veteran engineers by boiling down 15 pages of their hard work to three clean lines of code. Gosling persuaded him to work on the then-nascent Java project.

Van Hoff's talent, coupled with an entrepreneurial ambition, helped him build a tight-knit crew, each member hoping to get in on the interesting projects that always seem to bubble up around van Hoff. "Arthur is somebody who makes things happen," says Jonathan Payne, a Java veteran who, along with Adam Doppelt, has worked with van Hoff for more than six years. Though close, the three will scrap. "They're like old married couples," says Thomas Banahan, a former colleague of van Hoff's who is now global head of Lehman Bros. Venture Capital Group. "They all challenge each other, get grumpy, and go back to their corners." Not always. Pavni Diwanji, now chairman of spam fighter MailFrontier and a former Java team member, recalls a few fistfights during the all-nighters at Java central. "For Arthur, everything had to be perfect," she says. "There were no compromises with him." Still, when van Hoff left Sun in 1996 to help found Marimba, his team went with him.

That startup was one of the most hyped ventures of the pre-dotcom era. But van Hoff grew bored as an executive at the new company, which tried to develop ill-fated "push" technology to move Internet content to desktop computers. He preferred to code, rather than collate reports. "I'm not a manager," he admits. "I'm not interested in running companies. I'm much more interested in technology."

In 2002, van Hoff quit Marimba. Doppelt and Payne joined him in a new startup they called Strangeberry, a name the team came up with after a late night of random word association. Van Hoff didn't know what Strangeberry was going to do, but he started with two rules: Don't build anything like what Marimba was making. And have fun. (Van Hoff does have a lighter side; one of his hobbies is programming Lego robots to ride around on two wheels like a Segway.)

First the Strangeberry team built a jukebox programmed to run their music. But a lot of things the trio thought would be fun turned out to be "mind-numbingly dull," Doppelt says. Eventually they turned to brainstorming about the future of home networking, digital media, and home video, and how it would all converge in a house. Van Hoff was already a TiVo junkie, but he was also thinking way beyond the DVR. "We tried to imagine 10 years from now," van Hoff recalls. "What is going to be connecting the back of your TV?" Coaxial cable? Ethernet? Wireless? Ultra-wideband? The answer, he thinks, is irrelevant. "You won't be recording shows anymore," he says. "You'll just be viewing them. Why record them when you can just play them directly from the Internet?"

Van Hoff's crew built some basic DVRs and started experimenting. They dabbled with an intuitive interface that guided kids through the TiVo screen, and pitched it to the company, but TiVo wasn't interested. In July 2002, Payne bought a $300 AudioTron, a piece of hi-fi gear that transmits music files from a computer to a stereo; he'd ripped all his CDs but still wanted to listen to them on his stereo. Payne says his colleagues "all made fun of me" for paying so much instead of building something himself. But it set Doppelt thinking, and within days he pitched the "videotron." As he put it at the time, "Marry the power of the PC and the elegance of consumer electronics, and you have a killer product."

This was not a particularly startling revelation; people started working on the PC-TV meld many years ago. What was different was that van Hoff and his team had the coding chops to come up with a new software-based approach to convergence's central problem of getting various devices and forms of content to work easily together. The group dived into the project--and nearly blew apart. "There were some knock-down, drag-out fights" over technical directions and how to steer the business, Doppelt says. He considered quitting many times, but despite "butting heads" with van Hoff, he stayed. All the squabbling, Doppelt says, was ultimately good for the team: "Out of this crucible, you get something beautiful. If everyone is willing to get along, you get mediocrity."

Remarkably, the team had a prototype up and running within two weeks. By fall of 2002, they were showing VCs a box that could transport content to a TV screen from virtually anywhere. The VCs had two reactions, Doppelt recalls. The first was "We've never seen anything like this." The second was "How the hell is a software company going to make money in the consumer electronics business?" The VCs politely passed.

Six months later, van Hoff wangled a meeting with more VCs, this time at Redpoint Ventures. Ramsay, TiVo's CEO and an engineer himself, dropped in as a casual observer; Redpoint was one of his investors. He was floored. "We had a meeting of the minds," Ramsay recalls. The acquisition came together within six months. If it ever troubled van Hoff that TiVo was bleeding, he isn't saying. "We really liked TiVo," he says. "Great product. Good karma. Great engineers."

So how does Strangeberry give TiVo a fighting chance? For starters, van Hoff and his team joined up and brought their coding wizardry. Far more important, of course, is the Strangeberry software and what it promises for TiVo.

Right now TiVo machines can do their recording and time-shifting magic only on programs piped in from standard cable or satellite feeds. That has made it easy for cable and satellite operators to attack by, for instance, offering cut-rate digital video recording as a way to hold on to new customers. Time Warner Cable sells DVR service for $3 to $4 a month less than TiVo's $12.95 offering--and you don't have to buy the TiVo box, which runs about $150. Meanwhile, Microsoft and partners Gateway and Hewlett-Packard are blazing away from a different angle, trying to make the PC, not the television, the living room's hub. Add in the fact that some in the entertainment industry see TiVo as a tool for the potential Napsterization of film and it's easy to see why so many people have blown taps for TiVo.

The copyright issues will take years to sort out, but for the more immediate competitive pressures, Strangeberry provides several defenses. First, van Hoff and his coders are expected to quickly create new features for upcoming TiVo boxes that should help the company pump up subscriber revenue. Already, the Series2 TiVo enables users to move music and photos from computer to TV. And it doesn't hurt that more and more consumers want a DVR: The overall market is expected to hit 25 million units in 2008, up from 4.3 million this year.

The bigger payoff, as TiVo sees it, will come from far more elaborate things that van Hoff's software will enable. One of Strangeberry's distinguishing bits of technical hocus-pocus is that it can recognize any digital content format--MP3, HTML, DivX, and the like--and massage it into the familiar and friendly TiVo interface. That enables Strangeberry boxes to quickly and easily draw in anything on the Web and move it to the TV, which consumers generally prefer to the PC as a comfortable place to get their entertainment, especially video. The content then becomes subject to all the tricks--freeze-frame, ad zapping--that TiVo is already known for. When fully loaded TiVo boxes hit the market next year, they will add new layers of interactivity and potential customization.

Doppelt cites this example: Fox Entertainment could green-light an American Idol app that would allow viewers to vote with a click of the TV remote rather than by phone. Fans could zip back and review clips of the performers while making their decisions. Another possibility: TiVo users could swiftly tap into Netflix's vast catalog of films and enjoy them on-demand rather than waiting for them to arrive in the mail. (Netflix has said it will put its catalog online next year, and TiVo CEO Ramsay is a Netflix director.) Or ESPN could use Strangeberry technology to create customized channels--say, an A-Rod network that would send daily clips on the MVP shortstop's doings to a system's hard drive. "Strangeberry works and it's totally cool," says Marc Canter, co-founder of Web media tool maker Macromedia, who was given an early view of the technology.

None of TiVo's current competitors can do all the things Strangeberry makes possible. DVR devices built by Scientific-Atlanta, which makes set-top boxes for Comcast, Time Warner, and other cable companies, have recording capability but no Internet connectivity (see "The 800-Lb. Copycat," right). Microsoft has stumbled repeatedly with its MSN TV service; these days it's more focused on its Media Center software, which uses a computer screen, not a television.

Selling more of its own boxes won't save TiVo, and TiVo knows it. Its real bet is that the Strangeberry software's power and appeal will drive licensing revenue skyward. And the targets of TiVo's gamble are some of its biggest tormentors. Cable and satellite operators have spent millions vainly trying to create something that does what Strangeberry does; TiVo hopes to persuade them that it's cheaper and easier to simply license van Hoff's software. Ramsay says he's been talking to all the big cable and satellite operators; no luck so far. But David Miller, an analyst at research firm Sanders Morris Harris, says that is nonetheless TiVo's road to salvation. "I think TiVo will survive, but as a provider of technology to cable companies" and not a hardware maker or service provider. "The right question to ask," he says, "is whether they can make big money off licensing deals."

Another intriguing strategic twist involves the company's plans for ads. The feature that gave TiVo instant buzz when it debuted was its ability to skip commercials; many advertisers have loathed it ever since. But TiVo's technology collects reams of data on viewing habits, and its user base is heavily skewed toward the 18- to 34-year-olds whom marketers lust after. Ramsay believes Strangeberry software could create highly interactive ads. It will take time to get advertisers' trust, he concedes. But he says that within three years, basic digital recording, which today provides about 90 percent of TiVo's revenue, will drop to about a third. The rest will come mostly from software licensing and ads.

In the end, TiVo still could simply be overwhelmed by the pressures it faces--even if the Strangeberry software itself is a hit. Many observers think a likely outcome is that rather than partnering with rivals, TiVo will be swallowed by one, as happened to Netscape and many other tech pioneers that couldn't parlay their breakthroughs into lasting independent businesses. Van Hoff insists that this won't happen. "I'm at the center of a revolution," he says. If he succeeds, he'll probably lose the first part of the usual obscure-but-revered tag that people use to describe him. And he'll force a lot of hasty rewriting over on the obit desk.