The billionaire co-founder of Microsoft wants his elite research troop
-- unlike the corporate labs of IBM, Xerox and Microsoft -- to be free
from commercial pressures so they can pursue that valley rarity, pure
research. Only after four or five years will they have to start
spinning ideas into profitable start-ups.
This was 1992; the Internet and the World Wide Web as mass
communication tools were in their infancy. Allen felt the personal
computing revolution was stagnating. So his idealistic Palo Alto-based
lab, Interval Research Corp., would leverage this "interval" between
PCs and the next great wave, looking over the horizon and designing
technology that would have an impact in five to 10 years and drive the
computer industry for as long as 20 years.
That was the dream.
The story of Interval's demise is a cautionary tale not just for
researchers, but for all entrepreneurs and Silicon Valley dreamers,
revealing how the best brains, wads of money and abundant creative
freedom can implode, leaving behind a stack of unexploited patents,
accusations of betrayal, damaged careers and failed start-ups.
"They were plopped down in the middle of the greatest technology minds
on the planet, in the middle of the biggest revolution of the century,
and they never came out from behind their sand bags," scoffs Paul Saffo, director of the Institute for the Future.
"They hermetically sealed the place from Day One, and it meant they
lost touch with the intellectual life of the valley -- potential
collaborators and business partners."
Over the years, such critics had been skeptical that Interval would
ever produce groundbreaking work. The few projects that Interval
publicized heavily included a traveling exhibition of computer games,
two interactive toy start-ups and another one focused on computer games
for girls. After its death, the criticism intensified:
Did Interval researchers lack the speed, business savvy and risk-taking bravado needed to launch successful start-ups?
Or is it so simple? While Interval is being dismissed as a waste of
eight years and the work of great minds, there lingers this very
tantalizing thought: Maybe Allen has the keys to groundbreaking
technology tucked away inside the 140 patents that now are being
scrutinized by technologists at the Seattle headquarters of Allen's
investment firm, Vulcan Ventures, Inc.
Interval's collapse remains a valley mystery for this reason: Even in
death, the defunct lab is embalmed in secrecy. A Vulcan spokesperson
confirms that, in order to receive service-based severance pay,
employees had to sign what they refer to as gag orders. Says one former
staffer with a sarcastic laugh, "The joke is, 'We're so secret that no
one even knows we've died.' "
Let a thousand flowers bloom
Everyone agrees that, for the first few years, Interval was an
incredible place to work, a rare "let a thousand flowers bloom"
approach to research.
"I went to Interval because of the really interesting people. Some of
the ideas were out there, not the typical stuff floating around in the
computer science world," says Interval alum Sean White, former vice
president of technology for Lycos.
White signed on to Interval in 1992, fresh out of Stanford, where he'd
been a star computer science student. He'd never seen such a wild mix
of talents in a tech lab.
Inside a series of low, nondescript industrial park buildings at 1801
Page Mill Road, mechanical engineers, programmers and signal
computation experts collaborated with videographers, behavior
scientists, psychologists and visual artists, a group totaling about
150 full-time staffers and contractors at the lab's peak.
It was all part of the effort by David Liddle, Allen's hand-picked CEO,
to pioneer a new kind of research in which the focus was on what
consumers -- not businesses or technologists -- would want in the year
2002 and beyond. He and Allen were betting that computers would be
woven into the fabric of people's lives -- TVs, stereos, bedrooms,
kitchens, cars and clothing.
| |
 |
| |
Photo by Gary Reyes / Mercury News
Interval's CEO, David Liddle, with a 3D interactive kiosk developed at Interval. |
"We were exploring new media, devices and
services that individuals would choose to buy in a discretionary way,"
says Liddle, 55, an imposing professorial figure with gray hair and a
beard. "So it was important to understand early in the game aesthetics
and preferences. Technologists just aren't great at that."
Liddle should know. At the legendary Xerox Palo Alto Research Center (PARC), during the '70s, Liddle led the team working on the Xerox Star,
the first PC to use a graphical user interface -- easy-to-use words and
icons -- instead of text commands. But Xerox, worried that PCs would
undermine its copier business, did not commercialize the Star; Apple
and Microsoft seized the opportunity and launched the PC industry.
Liddle quit in frustration, moving on to found Metaphor Computer
Systems, a business software start-up that made him a millionaire when
IBM acquired it.
Allen and Liddle viewed Interval as the next PARC but with a twist: It
would be part think tank, part venture capitalist, part business
incubator. (The lab's location smacks of symbolism: Perched on a knoll
on Page Mill Road, Intervalites could almost look down on PARC half a
mile away.)
Instead of the typical scenario in which technologists engineer
products, then test them on consumers, Interval folks spent weeks in
the field videotaping and interviewing people about their use of high
technology.
To get out of their "skins," Liddle urged his crew to don masks and
role-play during "informances" (information performances) so they could
imagine what kind of products a grandmother, a punk rocker, or a sports
fan would want.
 |
|
| In a roleplaying exercise, an Interval group wears masks so their own identities don't interfere with the parts they're playing. |
|
It was this novel approach that led Liddle and computer user interface expert Brenda Laurel, one of his first hires, to ask this question: "What would it take to get little girls to put their hands on computers?"
For 22 years, Laurel had been researching, designing and producing
video games, primarily for boys, at Atari, Apple and Activision. She'd
earned a rep not only as an expert in how people interact with
computers but also as an independent, opinionated iconoclast.
Laurel's employers had nixed several game products she'd proposed for
girls, claiming that girls considered computers too geeky. But Liddle
agreed with Laurel that the subject could produce groundbreaking
research -- and open a potential $6 billion market. With Liddle's
blessing, she undertook an 18-month, multimillion-dollar study to look
at gender inequities, play patterns and why there were no successful
computer games for girls.
After interviewing experts and 1,100 kids, Laurel's team concluded that
girls found he-man superheroes boring. They wanted complex,
relationship-driven games.
Laurel's team developed prototype games for the 8-to-12-year-old-girl market, and by 1994 the games were getting a thumbs up.
The idea for Purple Moon, Interval's boldest start-up bet, was born.
Pointing fingers
Deconstructing the life and death of Interval is difficult because so many employees do not want to talk. Take Liddle, now a general partner with U.S. Venture Partners
in Menlo Park. He initially bristled at the idea of an interview about
Interval. Friends describe Liddle as having a protective big
brother-like relationship with Paul Allen. Three weeks later Liddle
agreed to discuss Interval, but certain topics -- including Allen --
were off-limits.
| |
PAUL ALLEN |
| |
 Net worth: According to Forbes, the Microsoft co-founder is the world's fourth richest person, with a net worth of $28 billion.
Personal: 47; lives in Seattle
Portfolio:
His investment group, Vulcan Ventures, owns the fourth largest cable
system in the United States. Among the more than 140 companies in which
he owns a stake: Microsoft, DreamWorks SKG (producer of music, TV and
film), ZDTV (tech-focused cable channel), Priceline.com, NorthPoint
Communications (provides high-speed data services).
Other holdings:
Allen owns the Portland Trailblazers and the Seattle Seahawks; he also
built Seattle's new $240 million Experience Music Project, a tribute to
his music hero, Jimi Hendrix.
-- T.O'B.
|
Allen himself refuses to discuss what skeptics
are billing as his $250 million escapade, a mere drop in the bucket for
his $20 billion dollar investment empire. Allen, whose personal net
worth totals $28 billion according to the latest Forbes magazine
survey, has a portfolio that reflects the convergence of his passions:
high tech, telecommunications, entertainment, sports and the latest -- extraterrestrial life.
| |
|
Vulcan Ventures president Bill Savoy | |
As for Interval, once the crown jewel of Vulcan, spokeswoman Susan
Pierson Brown explains that the media-shy mogul declines an interview
because he's "said all he has to say" on the topic. Which is about zip.
Weeks later, we finally were hooked up with Allen's business czar,
Vulcan president Bill Savoy.
In that intervening period, several former Interval employees,
frustrated that they have nothing to show publicly after years of work,
agreed to talk (both on and off the record) about what happened. What
they described sounds more like a Greek drama than the goings on at a
cutting-edge tech lab.
There is plenty of finger-pointing over who's ultimately to blame for
Interval's flame-out. One camp heaps criticism on Allen and his
Vulcanites, claiming they were too impatient to let Interval play out
its bets, aborting start-ups and changing the lab's mission seven years
into the experiment.
Another group sees David Liddle as a brilliant research leader but a
flawed manager whose preoccupation with secrecy and control, and
loyalty to Allen, slowed decision-making, leaving the staff without the
business alliances needed to launch winning start-ups.
Savoy questions the tight control Liddle exercised over the spinoffs.
"We probably should have brought in more outsiders earlier so we
weren't breathing our own air," he tells me. "But David felt like they
were his kids, and felt personally responsible for them."
One former Intervalite, who spoke off the record, compares his time at
the lab to being shanghaied on an island: "The owner and manager of the
island had absolute power so we always were in the position of
subjects."
Apartment cats
Interval was a snug cocoon, flush with money and creative freedom,
leading Liddle to warn against the danger that projects, especially
potential commercial spinouts, would become "apartment cats": pretty,
plump creatures incapable of surviving on their own.
| |
 |
| |
Interval-designed wearble computers |
By 1995, as Laurel geared up to launch Purple Moon, Interval was
buzzing with dozens of projects, from wearable computers to Webcam-like
devices. A visit to Buildings C and D at 1801 Page Mill Road was a walk
into science fiction.
Glenn Edens,
inventor of the first laptop,and now president of AT&T Strategic
Ventures, led Interval's wired Homeworks project. "We did a personal
video recorder in 1994, a replay TiVo-like machine," Edens recalls. TiVo,
a set-top device produced by TiVo Inc. of San Jose, allows viewers to
watch what they want on TV whenever they want it. But Interval didn't
develop the device -- someone else had a similar idea and built a
company around it.
Why? For the same reasons that researchers didn't spin out
Internet-related technology. Says one frustrated researcher, "We never
could take our big ideas and launch them into little devices that
perhaps could seed future industries."
Every commercial proposal had to meet the test of Interval's original
mission: Would it grow an entire new industry of products?
 | |
| In 1994, new gadgetry got test-driven in the hallways in Interval. | |
And even though Interval's goal was ultimately to commercialize its
discoveries, there was no formal structure for quickly recasting
researchers and their innovations into entrepreneurs and start-ups. A
full-time business development person wasn't hired until 1998, a year
and a half before Interval's death.
Consider the story of Sean White, who as early as 1993 was proposing work on several Internet-related technologies.
Before graduating from Stanford, he'd been researching how people would
use information on the Internet. This was a couple of years before the
Netscape browser helped to transform the Internet from a
government-backed network, largely for academics, into an information
tool for the masses.
With Liddle's approval, White continued exploring how people would
interact online through text, video and audio. But White and Roger
Meike, another Internet proponent, failed to convince Liddle and others
that the public was about to log on and that Interval should make the
Net and the Web a prime focus of research. "They saw it [the Net] as a
rarefied experience of academics that wouldn't make it to the public,"
White says.
Adds Edens, 'It's not that Interval missed opportunities in the
Internet. We just missed the Internet.'
Liddle insists he never dismissed the Net as a "fad," although that's
the word several staffers recall him using in the early '90s. They also
recall him swearing that he'd never launch a dot-com. "God no," Liddle
retorts, explaining his reasoning for rejecting many proposed Net
spinoffs. The former lab chief says many ideas were just too early for
the marketplace. Others failed to meet the lab's tough financial test,
or if they did, Interval wasn't built to compete on Internet time. Says
Liddle, "Had we started a different firm in 1992, it would have been
different story."
 | |
| A staff gathering in Interval's office, 1994. |
|
In 1994, White and another researcher thought they had a potentially
winning business idea: a company that would provide technology and
services such as setting up Web sites for those unfamiliar with the Net
-- which at the time was just about everybody.
As the youngest staffer, he couldn't do much when his idea was shot
down, ultimately by Liddle. "I think Paul and David still wanted to
figure out some sort of Next Big Thing. So this was going to be too
mundane," recounts White. Mundane maybe, but a similar idea made the
founders of USWeb multimillionaires. Says fellow Internet crusader
Meike, "Some of our friends got incredibly rich and we didn't. It's OK,
I'll get over it." Pause. Laugh. "I'm lying. I'll never get over it."
In retrospect, White shakes his head at the whole process. "We took six
months to put together a business plan. An entrepreneur would have
taken six days. We had researchers running marketing spreadsheets." He
chuckles at the memory. "Everybody there was creative and smart, but
were they marketers? Absolutely not."
Frustrated staffers began to wonder if Interval itself wasn't evolving
into a monster apartment cat or, as Edens puts it, "an industrial park
cougar." After more rejected Internet proposals, White finally left to
join WhoWhere, an Internet search engine startup; its acquisition by
Lycos made his fortune.
Girls go bust
In fairness, some of high tech's best minds missed the Internet in the
early days. But once they realized it wasn't going away, they rebooted.
Interval didn't.
Had Interval become too insular?
After his experience at PARC, Liddle kept Interval's work sealed under
a strict secrecy policy. Talking with the news media without clearance
was grounds for dismissal. Employees protested, especially academics
who argued that their reputations hinged on publishing their research.
Outsiders as well were hesitant to collaborate with Intervalites.
Liddle contends that every major industrial research lab had the same policy.
But Xerox PARC's chief scientist, John Seely Brown,
who ran PARC for years, says that wasn't the policy at PARC. He
believes that Interval's secrecy policy backfired. "If you run too
secretive a shop, you lose connection to the outside world," says Brown.
Intervalites say there was an even more fundamental problem at the lab: an inability to launch sustainable start-ups.
 | |
| Brenda
Laurel, who developed girls' software company Purple Moon at Interval.
In the background in Purple Moon main character Rockett. | |
In November 1996, Interval launched its first three start-ups amid a
blaze of publicity: Purple Moon; Carnelian Inc., an Internet software
maker; and ePlanet, an interactive toy company. The media loved the
politically correct Purple Moon and its attempt to fill the gap between
action heroes and Mattel's Barbie games. Purple Moon's CD-ROMs revolved
around the friendship adventures of Rockett, a spunky pre-teen.
"I considered Barbie a 'culture crime,' " says Laurel, who viewed
Barbie's empire as a pink ghetto, catering to the stereotypical beauty
and fashion dreams of young girls.
But behind the glowing headlines, Purple Moon was piling up debt as it
battled to grab sales away from arch-rival Barbie's fashion design
CD-ROMs. It was an unexpected fight because, according to Laurel, no
one at Interval or Vulcan knew that Mattel was entering the girls'
software market. "We didn't have the intelligence," concedes Purple
Moon's founder, who now works as a technology design and research
consultant with the Nielsen Norman Group.
Purple Moon had been kept secret from the valley, which meant Laurel
and others weren't picking up information about the competition from
the normal sources. "David knew it was a land mine not having VCs
involved, because a healthy company needs venture investments to make
it tough, but we had nothing legally protectable," says Laurel,
referring to their fear that someone would steal their idea for a girl
computer game company.
When Purple Moon finally did reach out, it was rebuffed. No investors
bit at what Laurel describes as the high valuation set by the Interval-
and Vulcan-dominated Purple Moon board. Instead of weaning itself from
Allen's seed money, Purple Moon remained almost entirely dependent on
Vulcan and Interval for its funding.
| |
 |
| |
Purple Moon's CD-ROMs revolved around the friendship adventures of Rockett, a spunky pre-teen.
|
By the summer of 1998, the Internet was eating into the CD-ROM market.
Purple Moon CEO Nancy Deyo and Laurel needed an infusion of cash to
reposition the company as an Internet-based business. (Purple Moon's
Web site was going gangbusters, sometimes getting more hits than
Disney's site.)
So the Purple Moon team was thrilled when Nickelodeon, the popular
kids' TV network, pitched a partnership deal -- equity and an
opportunity to collaborate on TV and online projects. It was just what
Deyo and Laurel believed was needed to turn Rockett and the company
into a national brand.
But when Nick came in with a low valuation of Purple Moon, the board
rejected the offer -- even though Interval's star spinoff was rapidly
burning through its cash. Vulcan's Bill Savoy, one of the votes, never
took the Nick offer seriously. "They wanted to be a partner, build some
TV stuff while we kept footing the bill," Savoy says.
What happened next still baffles Laurel. In the fall of 1998, just
months after rejecting the Nickelodeon offer, CEO Deyo was ordered to
sell the company for the best deal she could get. Savoy, who admits
he'd always been "a little dubious" about Purple Moon, couldn't imagine
how it would ever make money. "Even today, if you said you're building
an entertainment site for 8-to-13-year-old girls, there aren't enough
of them online to justify it," explains Allen's business czar.
It's no coincidence that the order to unload Purple Moon coincided with
a Vulcan-ordered audit of Interval: Vulcan clearly was losing its
patience with the lab's slow pace, fast cash burn rate and floundering
spinouts.
Arch enemy Mattel was the only company that nibbled. Sources familiar with the negotiations say Mattel made a verbal offer to acquire Purple Moon pending a review of the company.
But before a deal could be finalized, Savoy announced in February 1999
that Allen was killing his prize start-up. The board voted to put
Purple Moon in Chapter 7 bankruptcy, which meant creditors would be
paid only pennies on the dollar.
"Web hits were at an all-time high and we'd just implemented
e-commerce," says a still bitter Laurel, who can't understand why Allen
killed Purple Moon when there was an active deal on the table. Liddle
is mum about what happened; however, insiders say he pushed to win more
time for Purple Moon but couldn't budge Allen or Savoy. Savoy has no
regrets: "My recollection is that Mattel wanted to buy certain assets
but not the whole company."
Deyo and Laurel were furious that creditors wouldn't get paid.
According to Laurel, Purple Moon's CEO went back to Mattel and within
days sealed a deal for the toy maker to buy Purple Moon
in a Chapter 11 bankruptcy fire sale. Rockett now would be Mattel's
property and creditors would get their money. Even though Mattel has
not done much with the Purple Moon property, the irony that Barbie's
maker saved the day, allowing creditors to be paid, isn't lost on
Laurel. "Pink money saved my ass," says the woman who celebrated her
40th birthday by feeding Barbies into her trash compactor.
An empty rat hole?
"The very public debacle of Purple Moon was the point at which I wrote
off Interval as a lost cause," admits Institute for the Future director
Saffo. "It looked like it had a chance and, at the very least, they
could have gotten it to a safe harbor."
By the fall of '99, Purple Moon, Carnelian and ePlanet all were dead.
Like Laurel, Adi Gamon, the non-Intervalite CEO of Carnelian, questions
why Vulcan killed the start-up when it did: Carnelian was one month
away from its first beta testing, with such prominent customers as
Charles Schwab and Dell Computer lined up. "A venture capitalist never
would have shut it down at that point," he says.
Gamon is candid about his inability to recruit non-Vulcan investors, a
problem echoed by other Interval spinoff executives and confirmed by
venture capitalists. They were wary of partnering with one of richest
guys on the planet, who could, in the words of one former exec of an
Interval company, abruptly change his agenda and act "on whim, not hard
business decisions."
Liddle grows angry at criticism that Vulcan was run like a family
business, contending that the real problem wasn't Allen, but the
Internet frenzy. "The venture community lost interest in consumer
technology projects," he says.
In fact, Allen himself was shifting his investment focus. His
fast-growing empire certainly involved collaborations with other
investors. But the primary difference was that, unlike Interval, his
newer investments generally involved existing businesses with proven
track records. According to Savoy, a long-range research lab just
didn't fit with Allen's cable/broadband portfolio.
Ironically, by the time it spun out its last venture, Fantasma,
in February of this year, Interval seemed like it finally was getting
the hang of birthing independent companies that could fly on their own.
"Interval doesn't control the company or have anybody on the board,"
says Robert Aiello, CEO of Fantasma, the wireless home networking
venture, pointing out that non-Vulcan investors hold the majority stake.
But it was too late for Shangri-La.
The beginning of the end started with a terse news release last September in which Allen announced that Interval's focus now would be on quick turnaround projects for his broadband communication companies, not long-range research.
And the release stated that CEO Liddle was stepping down, although he'd
retain the title of chairman. Was Liddle elbowed out by the man who'd
hired him? "Absolutely not," he says, denying reports of a rift. While
he fully supported the lab's new direction, Liddle says he left because
he wanted to pursue his own interests.
Over the next seven months, a steady succession of brains headed out
the door, with the staff dwindling to about 50 by April 21, the Friday
when Savoy unexpectedly flew in from Seattle and called an employees'
meeting.
"He said it had become clear that the Internet was moving too fast for
a lab like Interval to keep up. It was the last day of Interval's
operation," recalls a former staffer. Savoy says the lab simply wasn't
pulling its weight in Allen's investment portfolio: "We tried to do big
things, to have the right vision but the world changed around us."
After their tumultuous time at Interval, only about 10 employees have
accepted Vulcan's offer to help launch a new broadband research venture.
The irony is that Interval's innovations are just starting to make
sense in our increasingly wired world. One example is technology
developed by Sean White and others, nicknamed the Web Phone, which
anticipated the issue of privacy in homes where the Net is always on
(think DSL). An electronic screen allows you to be seen or heard --
only to the degree you want. The Web Phone is among that stack of
patents now being reviewed by Allen's staff.
Meanwhile, Interval's biggest legacy may be the void it has left.
Silicon Valley's leading researchers worry about who now will chase the
seemingly wild ideas, who will pay geeks to dream up the stuff that
will drive the next New Economy. Valley lab chiefs have started meeting
informally. "We are discussing how to make sure that all the efforts to
commercialize technology doesn't suck away the brightest researchers,"
explains Bob Iannucci, vice president of corporate research for Compaq.
Researchers aren't the only ones worrying. Asks venture capitalist
Stewart Alsop, "If one of the world's richest guys has given up, then
who's going to throw money down the rat hole?"
The betting money is on a stepped-up alliance of private industry,
university and government funding -- not another billionaire.
Meanwhile, Alsop, a partner with New Enterprise Associates,
has a more immediate problem. He'd love to recruit some Intervalites
for his companies. "My CEOs are trying to find people. But Interval was
so secret," he says with a laugh, "it's hard to know who was there."
TIA O'BRIEN is a Mercury News staff writer. She wrote this article for SV, the Mercury News Sunday magazine.
SV News |
In Time For... |
Columns |
Companies |
Investing |
Personal Technology
