Jamie Zawinski was fried.
He had fallen into the habit of working two days straight, sleeping for six hours, then beginning another two-day shift. During the time he was at the office, which was most of the time, he was focusing his time and energy on fixing bugs generated by his fellow programmers at Netscape Communications Corporation.
Painful, but no big deal. Except they were becoming a bigger deal by the day. The company was just weeks away from shipping the first commercial version of its software for the Internet, the Netscape Navigator. A week after its release, a million people might be using Zawinski's code to browse the World Wide Web's skyrocketing assortment of data files, audio clips, and digital images. Every moment counted; every detail mattered.
Suddenly, without warning, Zawinski's computer decided to reboot. All his pending fixes and unsent e-mails -- an electronic record of hours of drudgery punctuated by flashes of inspiration -- were vaporized.
He pounded his fists and knocked over his computer.
His chair, now a four-legged projectile, flew across the room. He put on his coat and left.
"I was pretty sure I had just quit," said the young programmer, whose goatee, partially shaven head with braids-cum-dreadlocks, and uncensored outlook on life made him something of a hero to his coding colleagues. ("We all enter this world in the same way: naked; screaming; soaked in blood," reads the epigraph that begins Zawinski's personal home page on the Web. "But if you live your life right, that kind of thing doesn't have to stop there.")
A few days later, Zawinski was back. Back poring over the thousands of lines of code that comprised his project -- a version of the Navigator aimed at the power tool of choice for Net junkies: the Unix workstation. Only three pieces of evidence offered any trace of his tantrum: the word "Angry" scrawled on a business card that listed his job title as Hacker; the hastily assembled effigy (a computer case covered with a blanket, a set of headphones sitting on top) that occupied his empty chair; and a quick addition to a collection of unofficial T-shirts that functioned as a time line of life at Netscape.
The "chair with wings" T-shirt, if nothing else, served as a symbol of intensity in action. For Zawinski and his fellow programmers, intensity was everything.
This was the fall of 1994 -- less than eight months after Netscape took shape, more than eight months before it would go public in one of the most frenzied initial public offerings in history. Today that IPO is the stuff of business legend. It valued the young company, with just $16 million in revenues, at more than $2 billion. It left Netscape's 24-year-old cofounder, Marc Andreessen, holding shares worth nearly $60 million and its other cofounder, computer veteran Jim Clark, 51, with shares worth half a billion dollars. It reserved a pool of 4.5 million shares for 26-year-old Jamie Zawinski and the other rank-and-file Netscapers -- a pool worth about $250 million. More than just creating wealth, the IPO also captured the popular imagination. Virtually overnight, Netscape was perceived as the defining company of the Age of the Web.
Back in 1994, as he decided to return to work on the Navigator for Unix, Jamie Zawinski wasn't thinking about defining an era. He and his colleagues were thinking about survival. Sure Netscape had one of the best pedigrees in Silicon Valley history. Its first $4 million came from Clark, the Stanford University professor turned entrepreneur whose first business creation, Silicon Graphics, makes the superfast visual workstations that have revolutionized everything from industrial design to filmmaking. Another $4 million came from Kleiner Perkins Caufield & Byers, one of the most unimpeachable names in high-tech venture capital. But even the best pedigree can't pay the bills, and in the fall of 1994, Netscape was down to its last million dollars. It was running out of money.
Worse, it was running out of time. Soon after they created their company, Clark and Andreessen had assembled ten programmers, nearly all of them in their twenties, six of them friends of Andreessen from college. To these young programmers, Clark and Andreessen presented a life-or-death challenge: obliterate Mosaic, the very product that brought the two founders together in the first place.
Marc Andreessen was an undergraduate at the University of Illinois, writing code for $6.85 an hour at the National Center for Supercomputing Applications, when he and a fellow student, Eric Bina, became intrigued by the potential of the World Wide Web, a new technology for linking the resources of the Internet. But the Web lacked a rich graphical interface -- an intuitive way for people to unearth the vast material it stored. In a wild burst of coding in the winter of 1993, Andreessen and Bina wrote the basics of a graphical Web browser called Mosaic. Almost singlehandedly, their work turned the Web into the business and pop-culture phenomenon it is today. They had created The Next Big Thing.
Now Clark and Andreessen wanted to turn The Next Big Thing into The Next Big Business. But successfully launching the Netscape Navigator meant overthrowing its predecessor. Hence their new software's code name: Mozilla, the monster that would slay Mosaic.
The good news, according to Jon Mittelhauser, 25, one of the original refugees who left Illinois to join Clark and Andreessen, was that Mosaic "really wasn't a very good first-generation browser." For all its revolutionary virtues, the program was slow, lacked many basic security features, and did not allow for especially rich document layouts.
The bad news is that technical merit seldom determines who wins and loses the competitive race. Being first is more important than being best. And Mosaic was first. By the fall of 1994, it had become a fundamental tool for as many as 3 million users of the Web; it was adding up to 600,000 users a month. Every week that went by made slaying Mosaic that much more daunting.
How to focus the energy of the troops on this all-or-nothing mission? Forget stock options, urged Andreessen, the group's technical and spiritual leader. Let's get personal. Andreessen pledged that if his colleagues could deliver the first commercial version of the Netscape Navigator by the end of 1994, less than one month away, he would do two of three things -- none of which would come easily to this stocky, slightly pasty, aficionado of fast cars and Pepperidge Farm cookies. One, he would strap on roller blades. Two, he would wear Spandex. Three, he would eat health food.
Hardly the stuff from which great leaps forward come. Except at a company full of 24-hour-a-day programming machines who ate whatever they could scoop from garbage cans stuffed with peanut butter cups and cereal; who blew off steam with midday games of roller hockey; and who alternated between amusement and anxiety at the pressures under which they operated.
November turned to December. Netscape turned from an office to a dormitory. Programmers crashed in their cubicles -- or, if they needed "real" rest, in the futon room. Jamie Zawinski's 130-hour weeks became the norm. Mike Barbarino, who worked on the Windows version of the Navigator (and who, now 35, is an old man by Netscape standards), spent 7 days a week at the office and slept on a futon for only 3 or 4 hours a night. On one of the few nights he was actually at home, programmer Lou Montulli, 24, received a call from a survey company researching work habits. When he told the researcher how many hours he had been working (110 to 120 per week), the worried researcher responded that his computer would not allow him to enter a figure that high.
But the never-let-up atmosphere paid off. On December 15, Netscape shipped the first commercial version of the Navigator. It was pay-up time for Andreessen. He arrived at an all-hands meeting in the company cafeteria, ankles wobbling, a pair of roller blades on his feet, a pair of acrylic shorts hugging his buttocks. Then, with great fanfare, he downed a healthy helping of tofu. Three for three.
Within four months, the dimensions of Netscape's win became apparent. With no advertising -- with no sales in retail outlets -- a stunning 6 million copies of the Netscape Navigator were already in use. Users simply downloaded the software directly from the Internet. Netscape was calling the shots.
Mosaic, if not already dead, was mortally wounded. Back in the fall of 1994, while the Netscape crew was still furiously writing code, visitors using the Mosaic browser accounted for 60% of all the traffic on the Web. By the spring of 1995, the Netscape Navigator accounted for more than 75% of all Web visitors. Mosaic accounted for just 5%. "If we had been six months later," Jim Clark mused, "we would have been lost in noise." Instead, by delivering on time, Netscape created a new browser standard for the Web.
Mozilla ruled. The competitive landscape had been redrawn. Jamie Zawinski could get a decent night's sleep.
Establishing a standard is different from establishing a brand. Launching a phenomenon is different from launching a product. Producing a movement is different from producing a profit.
You do it in a different way and a different time -- you do it in Netscape Time.
Netscape Time is only partly about speed, although it is most certainly about that. It's also a genetic endowment, an operating system cooked into the DNA of hungry young programmers going about their work. It is as much a mind-set as a business model. Part paranoid, part predator, it shapes everything Netscape does. It's hardwired into how a company in overdrive -- a company whose headcount, in 15 months, has gone from 2 to 330 with little sign of slowing down -- recruits and evaluates talent. It shapes its uniquely interactive relationship with customers. And it explains its coevolutionary relationship with the technology itself -- how Netscape uses the Web to win control of the Web.
In this bifurcated economy, where tired, sclerotic organizations struggle against long odds to cross the gap between those who get it and those who don't, Netscape Time is the defining birthright of a company born on the right side of the Great Divide. Indeed, in an economy where even breakthrough technologies become obsolete within a few years, where even the deadliest competitors must change their game in the face of changing circumstances, Netscape Time may be the company's most enduring invention.
These are its core principles.
Fast Enough Never Is
It was early February 1994 when Jim Clark learned that young programming whiz Marc Andreessen had moved to Silicon Valley. Almost immediately, Clark sent Andreessen an e-mail and asked to meet. The two hit it off; Clark invited Andreessen to devise a plan that would make waves; Andreessen proposed his Mosaic killer; the new partners hopped a plane to Champaign-Urbana and hired a bunch of Andreessen's college buddies on the spot.
By April, programming on Mozilla was moving toward full tilt. It took the new team just six months to release a beta version of the software that would overthrow Andreessen's original creation.
Ever since that initial Clark-Andreessen meeting, Netscape has maintained this lightning pace. Engineering Vice President Rick Schell says it's not unusual for the company to contact a hot young programmer on Monday, do a series of interviews on Tuesday, and say "welcome aboard" on Wednesday.
Netscape can move quickly because it knows what it wants. It targets programmers from best-of-breed schools (Stanford, Illinois) and speed-driven companies (Oracle, Silicon Graphics, Next). More than just brains, Netscape wants buzz -- programmers comfortable with the company's code-writing culture. Output is valued in the extreme. Activity itself means nothing. Fast enough never is.
The Paranoid Predator
The first job of Andreessen's team was to kill the product many of them had worked to create. Little wonder, then, that Netscape, having killed in order to be born, never forgets how quickly the predator can become the prey. This is a company that's always looking over its shoulder.
At least that's what Jim Clark hopes. His job, he says, is creating "paranoia" among the troops. So what if Clark himself is one of Silicon Valley's fabled entrepreneurs? So what if Netscape's senior management team includes celebrated executives such as James Barksdale, formerly a top player at both FedEx and McCaw Cellular? So what if Netscape's IPO is the envy of Wall Street? Clark's role is to undermine the glowing publicity, subvert the evidence of success, and instill fear and urgency at all levels.
In meetings of his senior staff, or at all-hands sessions in the company cafeteria, Clark portrays Netscape as a 300-person underdog up against the likes of mighty Microsoft and Oracle. And not without reason. Microsoft's new online service, The Microsoft Network, comes equipped with its own Web browser -- based, ironically, on Mosaic, and licensed from Netscape competitor Spyglass Inc. If just 10% of the projected 40 million users of Windows 95 choose to access the Web through Microsoft's Internet Explorer, the game changes -- radically. Add a collection of other rivals - some of them brand names, many of them well-funded -- who are building, licensing, and otherwise upgrading Web browsers, and paranoia looks more like justifiable anxiety.
Especially in light of the stakes for which Netscape is playing. Netscape makes no bones about its strategic intent. It is not out merely to prosper. It is out to dominate. It wants to be the Microsoft of the Internet. Anything less will not be just disappointing; it could be fatal. Which is why Netscape's leaders do not see the company's 75% market share as a competitive aberration. It is the precondition for success.
Skeptics continue to wonder about Netscape's strategy: How can a company that aspires to dominate a market give away its core product? In fact the Navigator, while certainly Netscape's most famous product, is not its core product. The Navigator is the market-maker by which Netscape establishes a standard. Its growing collection of server products -- complex software that companies use to post information on the Web and conduct electronic commerce --are the revenue generators through which Netscape will earn the bulk of its profits.
"Netscape builds printing presses," says President James Barksdale. "But first I've got to teach everybody to read, or there won't be any publishers."
Jim Clark offers a simpler explanation. "This is not freeware," he says, "this is marketware."
All Work, All The Time
How does a predator avoid becoming prey? How does a company with 75% of its market increase share against well-funded rivals? It's simple: nobody ever stops working.
General Manager Jim Sha says he routinely spends 11 hours a day at the office, joins his family for dinner, then works late into the night. But there are still not enough hours to go around. His all-too-common lament: "Where do I find another 10 hours" a week? So he and his colleagues improvise. Stuck in a meeting? Check e-mail on your laptop. On your way to a stress-alleviating touch football game? Debate alternative approaches for painting graphics on a computer screen. Eager to see your wife and kids? Buy computers with videoconferencing capability and visit while you're still at the office. This last wrinkle, from Mike Barbarino, never got past the wish-list stage. But in a few months? "Nobody's really pushing us to do this," Barbarino explains. "We tend to do it to ourselves."
Not long ago, for example, Barbarino learned that Netscape's biggest customer, MCI Communications, needed a new feature for its customized version of the Navigator. It was a feature, amazingly enough, that he and his fellow engineers had failed to anticipate: advertising. Overnight, Barbarino created the necessary hooks -- pieces of programming code that allowed MCI to slip advertising into designated spots -- and met a non-negotiable deadline for a prototype. So it goes when you're working in Netscape Time.
"Just Enough Management"
Who makes Mike Barbarino labor all night to meet a deadline for MCI? He does. The only people who work this hard are people who want to. The only people who want to are people with enough freedom to do the work they want to do.
Netscape is a company that consciously undermanages. Need proof? Consider the roles of Clark and Andreessen, the two names synonymous with the company's meteoric rise. Neither, it turns out, plays a central management role at Netscape. Clark represents the company to its investors and strategic partners. He says he's at his best as a manager in organizations with fewer than 100 people, and Netscape crossed that threshold long ago.
Andreessen, meanwhile, no longer writes code. Nor does he directly supervise programmers. Instead, he's Netscape's technical visionary and Internet ambassador. He is still a huge presence at the company -- popping in and out of meetings, critiquing code-writing priorities, dreaming up new features for the software, fielding up to 200 e-mail messages a day. But he's not a manager in the conventional sense.
Formal "management" falls to a trio of veteran high-tech executives: 52-year-old President and CEO James Barksdale; 45-year-old engineering chief Rick Schell (Symantec, Borland); 45-year-old general manager James Sha (Oracle, Wyse Technology). All three are clear that their major responsibilities are to set priorities and to establish schedules -- and to stay out of the way.
"Just enough management" is how Andreessen describes the philosophy. "If you overmanage software," he says, "the result is paralysis." Schell compares it to "herding kittens." Tell programmers what to do and they are guaranteed to rebel. And the first casualties of rebellion are productivity and creativity.
The heart of this company -- the people who make Netscape Time tick -- are the rows of young programmers who sit for days at a stretch, absorbed by their work, suspicious of the suits. Programmers like Mike Holst, who boasts that he has "avoided management like the plague" throughout his career. Or Lou Montulli, a Netscape star whose creations include the Amazing Fish Cam, an utterly useless (and thus popular) device through which Web surfers can visit the aquarium in his office. Spiritually at least, Netscape is a place where the inmates run the asylum.
Four Times Faster
Engineering chief Rick Schell gives the ultimate definition of Netscape Time -- turning out new product releases four times faster than the competition. He knows it because he's lived it. Netscape Navigator 1.0, released last December, was designed to run on the Macintosh, Windows, and X Window machines for Unix. Less than three months later, Netscape released Navigator 1.1. This version added a bunch of new features and ran on 32-bit platforms such as Windows NT and the Macintosh PowerPC. Three months later, Netscape released a beta version of Navigator 1.2 designed to run on Windows 95 and announced plans to release the official version two months after that.
And that's just the Navigator. Servers, which range in price from $1,495 to $50,000, roll out the door just as quickly. The Netscape Communications Server is the baseline product with which companies publish hypermedia documents on the Web. The Netscape News Server allows companies to create their own private and public discussion groups. The Netscape Commerce Server allows companies to conduct electronic transactions over the Internet. All told, in its first 15 months, Netscape announced or shipped 11 major new products.
How fast is fast? Clock it against the track records of some other famous names in software. It took Kaleida Labs, the much-ballyhooed joint venture between Apple and IBM, three years to release its ScriptX multimedia programming language. For the last five years, General Magic, another celebrated Apple offspring, has been building a market for its operating system for handheld communicators. Windows 95 was first conceived in 1991 and code-named Windows 93. Four years and countless delays later, Microsoft's new operating system finally hit the market.
To Netscape, these are slow companies. "I don't know anyone that's made it going after a five-year dream," Jim Clark says. Which explains why plans for even the most ambitious new products are not allowed deadlines longer than six months. Time is everything.
"The only competitive advantage is speed," says Jim Sha, the Netscape general manager responsible for integrated applications.
"When you live on the Net," adds Barksdale, "you realize there is very little excuse for delay."
Is the web a technology? Or is the Web a place? Is the Web an experience? Or is the Web a relationship?
Netscape understands the Web so well that it has positioned the Web at the heart of how it operates. Netscape is both on the Web and of the Web. It uses Web software to win control of the Web. It then uses its presence on the Web to create new-and-improved software that extends its control. It then uses that software to ... well, you get the idea. There would be no Netscape Time without Web2.
The Web squared approach begins with a mantra championed by Marc Andreessen: "Worse is better." In other words, it's more effective to release usable software quickly than to wait for perfection -- especially if Netscape can use the Web to access the source of perfection: the market itself. Worse is better because worse is faster.
"Doing it `right' is a time-sink," says Andreessen, who urges his programmers to revise their code early and often. "You get closer to the market by using the market," adds Deepak Puri, Netscape's director of business development.
Thus, Web squared. Yes, the Web is a market. But the Web is also a feedback loop, an electronic channel between producer and customer that allows for nearly instantaneous interactions. So Netscape engineers don't develop software; they codevelop it with their customers. The product is the process is the product.
Each new version of the Navigator is released online rather than through retail outlets or resellers. An interested user can download an "evaluation copy" at no charge and later decide whether to pay $39 for a fully supported version of the software. The entire process takes about two minutes. Interaction with customers -- and thus, the design of the next version -- can begin two minutes after the download, once the feedback rolls in.
And feedback does roll in. Netscape's home page -- the first stop every time someone launches the Navigator -- provides an electronic Bug Report Form through which users can identify problems. Users can report the problem's severity; how often they can reproduce it; where a Netscape engineer can visit (electronically, of course) to see the problem firsthand. The bug form also invites users to identify features they want in future versions of the software. There's plenty of feedback because the process is direct. All it takes is the click of a button.
Bug Report Forms allow one-way communication; Netscape's engineers are so swamped with feedback they can't respond to each message. But the Web also permits two-way interactions. Netscape's home page lets users participate in a collection of Netscape User Groups -- NUGgies in company parlance. NUGgies are electronic forums where customers talk to other customers and to the company's engineers; Netscape engineers discuss new features and priorities; newbies and veterans swap questions and war stories.
It's a free-for-all of ideas -- with all the ideas dedicated to upgrading the company's products as rapidly as possible. There are forums devoted to six different product categories and issues -- from the Navigator to Servers to Internet security. Each forum is organized into a set of folders on problems and features. Each folder contains in-depth exchanges between users and engineers on everything from the design of the Navigator's "bookmarks" to the speed and reliability of a particular server.
NUGgies are Netscape's rolling ballot box. To get a quick read on what customers are most passionate about, engineers only have to rank the folders by the number of comments they contain. How important is this web of communication to effective product development? Says marketing vice president Michael Homer, "In principle it's the whole story."
But that story couldn't be written if Netscape operated like a "normal" software company -- that is, if it operated off the Web. If Netscape relied on standard retail distribution, the physical acts of manufacturing disks, shipping them across the country, advertising their arrival, and waiting for customers to make their purchases would take months rather than minutes. If Netscape relied on traditional market surveys, the process of mailing disks to testers, following up with questionnaires, and waiting for and sorting through returns, would take months more. At Netscape, working off the Web translates into unthinkable delays.
For most companies, unthinkable delays are called business-as-usual. Consider Microsoft's August 24 launch of Windows 95. Microsoft was closing in on final code for its new operating system much earlier in the year. So why did the release wait until late summer? For one thing, Microsoft spent months distributing 400,000 beta copies so users could test for bugs. Still, it achieved final code on July 14. Why wait six more weeks? Because Microsoft had to organize a dozen manufacturing plants and 500 trucks to produce and deliver the software to 20,000 retail outlets. Life off the Web is awfully messy.
There are, to be sure, risks to life on the Web. To an extraordinary degree, Netscape has opened itself up to the competition. Engineers from Spyglass, America Online, or any other company can log on to Netscape's user groups and see what its customers are saying, what its engineers are promising, what glitches are raising a ruckus. It's as if Pepsi published the results of its taste tests in a public forum that Coca-Cola could visit every day.
Netscape's executives don't seem worried. The company doesn't post its official timetables and development priorities. Those plans are discussed and executed on Netscape's internal Web site, its Intranet. Besides, there's something satisfying about watching the competition surf for intelligence on Netscape's own site -- especially since they can't surf it without using Netscape software. (It's the only browser that can "read" the user groups.)
Marketing Vice President Homer relishes the thought: "The competition has to use the Netscape Navigator to learn about the Navigator. The good news is they can download it for free."
Tom Steinert-Threlkeld (firstname.lastname@example.org) is Editor-in-Chief of "Inter@ctive Week."