Mitchell Baker and the Firefox Paradox
"Whoever just joined, you're breaking things up. It might help if you muted your phone."
"We got a lot of good stuff done last week."
"Fish got that working; that's pretty cool."
"We need to get serious."
"Whoever is talking right now has great ideas."
"Please send blankets."
Welcome to the weekly meeting at Mozilla Corp. It's a bit hard to say who's in attendance because it's open to everyone. That's "everyone," as in the human race. The phone number and online chat coordinates are publicly posted, as is the date, time, and agenda, for anyone who might feel like having a hand in Mozilla's strategies and operations.
The comments fly, seemingly at random or at cross-purposes. Mentioned almost in passing is word of the completion of the software program that has been released to the public as Firefox 2. Firefox, of course, is the Web browser that has established itself as the one serious competitor to Microsoft's (NASDAQ:MSFT) utterly dominant Internet Explorer. It has been downloaded more than 200 million times, making it one of the hottest high-tech products in history. It is so hot, in fact, that Microsoft has been forced to play catch-up and release its first significantly retooled version of Internet Explorer in five years, which offers some of the features that Firefox has long boasted. One can only imagine how difficult it would have been to get a ticket to the meeting at which Microsoft managers reported its completion.
The Mozilla meeting quickly moves from software progress reports, which don't seem to ignite much interest, to a discussion of what Mozilla can do to create closer ties to the thousands of volunteer programmers around the world who are largely responsible for developing Firefox and Mozilla's other, lesser-known products. This subject generates a lively discussion, and draws several comments, offered in a distinctive verbal style that manages to be laconic, forceful, and a bit playful all at once, from a heretofore mostly silent participant. When someone wonders whether Mozilla should offer an organized program to help train volunteer developers, this voice quickly chimes in: "Maybe we can find a way to have smaller groups that self-organize and then get together." And then, after a moment, the voice suggests essentially the opposite. "Maybe we should hire somebody here to be the focal point."
The voice belongs to Mitchell Baker, who leads Mozilla under the official title of Chief Lizard Wrangler. That Baker is fond of tossing out what seem like contradictory ideas in quick succession is entirely in keeping with a background that includes being a California Caucasian who speaks Mandarin, a political lefty who ended up at a corporate law firm before taking over a fiercely anticorporate organization, and a devoted parent with a passion for performing on the trapeze. Baker even seems to have two different haircuts that face off on opposite sides of her head, unified only partially by a dose of red dye. Finally, in spite of the first name, Baker is a woman, one of the few who have risen to prominence in one of the most male-dominated segments of the high-tech industry.
As a mass of apparent contradictions, Baker, 48, is a good fit for Mozilla--which is, after all, the profit-making arm of a nonprofit organization, a community-run company whose executives cut secret deals with big businesses, a 70-employee start-up that has threatened the ambitions of one of the world's corporate titans, and a well-funded company that depends mostly on unpaid enthusiasts to develop its software and to handle its marketing, customer support, even strategic planning. As whimsical as the title Lizard Wrangler may be, it's probably closer to the mark than "CEO" or any other term the conventional business world has to offer. "Baker and Mozilla are wrestling with new questions about the boundaries between communities and the corporation," says Siobhan O'Mahony, an assistant professor at Harvard Business School who studies new forms of organization. "She's creating a new management model based on engaging a business ecosystem." Baker, and the organization she leads, clearly are not much like what has come before them. But they may be a lot like what entrepreneurs and companies will become in the years ahead.
That Mozilla and Mitchell Baker have been heralded as the vanguard seems highly unlikely, especially considering that both came to prominence amid the excesses and bumbles of the dot-com boom and bust--specifically the browser wars. By 1998, Netscape Communications had seen its Navigator browser, the software that opened up the Web to most of the world, yield its dominance to Microsoft's newer, faster Internet Explorer. In an effort to broaden its impact, Netscape had embedded Navigator in a small suite of e-mail and other programs called Communicator, but the result was bloated and clunky. In a drive to compete against Microsoft and salvage its fading reputation among the tech-savvy, Netscape made Communicator an "open-source" product--that is, it publicly released the programming code so that anyone could tinker with it. In 1999, the company spun off and funded a small project called Mozilla (the name supposedly was derived from "Mosaic killer," Mosaic having been the first browser) to coordinate the tinkering so that there would be an official product, albeit one that wasn't controlled by Netscape.
Baker, then a Netscape lawyer, was assigned the tricky job of writing a software license for Mozilla that would permit people to alter the program without allowing them to convert the results into a proprietary product. She proved so adept at finding common ground in the often intensely conflicting needs and styles of her corporate employer, the tech-obsessed and sometimes militant open-source community, and the world of users, that Netscape asked her to run the Mozilla project. She might reasonably have passed on that opportunity, given that almost everyone expected Mozilla to fail. But Baker found the offer irresistible.
She recognized at once that Mozilla was a chance to help shape a new kind of organization that existed outside the bounds of corporate governance and of many of the ordinary rules of work. Most of the contributors would be volunteers, and the coin of the realm would be not salary or title, but respect, accomplishment, camaraderie, and challenge. It would not be an uncoordinated free-for-all; the community, for the most part, would need to agree on the direction the project would take and on who would be given responsibility for a task. Mozilla would be a meritocracy. If you proved talented and diligent, you'd get more important tasks and ultimately acquire some level of project leadership, sidestepping much of the politics and bias of traditional corporations.
It would be easy to assume that leadership is less important in this sort of community-driven organization. In fact, the opposite is true, says Sandeep Krishnamurthy, who is an associate professor in the business administration program at the University of Washington, Bothell, and has studied Mozilla. "Someone has to take the lead and reach out to this large group of people to provide feedback and motivate them," he says. In some ways, he adds, the bar at an open-source organization is higher for a manager, not lower. Unlike employees, volunteers generally won't put up with inept, bullying, or unfair managers. They'll just walk away. Edicts won't work when it comes to getting this sort of a community moving in a common direction. Rather, it takes a combination of inspiration and persuasion to build consensus.
If Baker proved an effective leader, it would be in spite of her not having the two most common characteristics of open-source movers and shakers: being a man and being a programming whiz. It's estimated that less than 2 percent of the open-source community is female. Meanwhile, the percentage of respected open-source leaders who didn't get to their positions via daring feats of coding is probably just as small. "Just because it's open source doesn't mean it's open door," says Krishnamurthy. "To get anywhere, you have to win the respect of an elite group of people who develop code." To do that entirely through nontechnical management skills would be nearly unprecedented, to do it as a woman all the more so.
But that was not the first challenge Baker faced. Soon after she took the helm of Mozilla, AOL (NYSE:TWX) agreed to buy Netscape for $4.2 billion, a deal completed in 1999. The work on Mozilla continued. But lacking a clear prospect of a fast return on its investment in Netscape, AOL started to clamp down on the project's costs, laying off Baker in 2001. But while AOL could cut off Baker's salary, Mozilla was an independent entity, and Baker, who had become a popular and respected figure in the open-source world, remained as an unpaid volunteer for about a year, until a nonprofit called the Open Source Applications Foundation offered to restore a portion of her former salary to support her Mozilla work.
In 2002, the Mozilla project released its first official product, Mozilla 1.0. It was a suite of Internet applications that integrated a Web browser with programs for e-mail, online chat, bulletin boards, and building websites. The program worked well, but by this time the world had largely accepted Internet Explorer, and there were other simple programs for e-mail and other tasks. Few computer users wanted to start all over with a new, relatively complex piece of software. Mozilla 1.0 was the solution to a problem that no one seemed to have.
That might have been the end of the story, but it turned out there was a small side project taking place within Mozilla that until then had received relatively little attention. Two young programmers, Blake Ross and David Hyatt, had been working on the browser portion of Mozilla 1.0, breaking it down and reassembling it in leaner form. They were like two guys in a minivan factory dragging parts off the assembly line to a dark corner of the building to assemble a dune buggy. The result was a simple, speedy browser. Now, in 2002, with Mozilla 1.0 failing to cause a stir, a question presented itself: Was it possible that while most of the team had been toiling away on a doomed Internet suite, Ross and Hyatt had quietly thrown together the basis of an IE killer?
It was the right question, and the right time to ask it. By 2002 Internet Explorer's real shortcoming was starting to show itself: security. In order to make the program work with a wide range of Web technologies, Microsoft had made it easy for website developers to get access to the guts of the program while it ran on a user's machine--which meant the IE user's software could also be exploited by hackers, spammers, and other unscrupulous Web bottom feeders. The public, and corporate networks, became plagued with viruses, endless pop-up windows, and spyware. Baker understood that Mozilla suddenly had a new mission: Make browsing safer.
The new browser was released as Phoenix in late 2002. In an effort to keep the browser simple and secure, the core program provided only the basics in viewing a website. But it also made it easy for programmers to write "extensions" that would add other features--anything from built-in dictionaries to at-your-fingertips weather reports, which could be added to the browser with a few clicks, enabling users to make their own decisions about how to balance simplicity and strength. By early 2003, Phoenix was starting to attract attention among the tech-savvy Web avant-garde. Nevertheless, AOL signed a seven-year agreement with Microsoft to make IE its default browser. It also laid off all the Netscape programmers working on Mozilla.
Baker started firing off e-mails to her growing fan base of computer-industry leaders, many of whom were concerned about Microsoft's de facto hegemony in Web browsing. If everyone used Explorer, Microsoft would be in a position to dictate technical standards for websites and could then in theory integrate the resulting proprietary approach to Web browsing with its Windows and Office products--a strategy toward which Microsoft appeared to be edging, and that threatened to close off much of the Web to users of Linux, Apple, and other non-Microsoft software. Creating a popular alternative to Explorer would, in effect, short-circuit any such Microsoft effort. Lotus founder Mitch Kapor chipped in $300,000, and IBM (NYSE:IBM), Sun Microsystems (NASDAQ:SUNW), and Red Hat (NYSE:RHT) offered the services of dozens of programmers. Baker, ever the persuasive diplomat, was careful not to burn any bridges with AOL; the company even agreed to donate $2 million to the cause. Baker spearheaded the formation of a formal nonprofit corporation called the Mozilla Foundation to be the official overseer of the code and funding.
In 2004, Mozilla was ready to release a preview of the latest version of its browser, now called Firefox. The timing seemed propitious. The entire PC universe was up in arms about the exploding problem of Web-based malware. IE was considered so vulnerable to hackers that the computer security arm of the U.S. Department of Homeland Security recommended dumping the application, and some private computer security firms echoed that advice. Firefox, by contrast, was simpler and designed specifically to resist hackers and viruses. (Microsoft declined to respond to questions for this article. "We certainly respect that some customers will choose alternative browsers," the company said in a written statement.)
Finally, Mozilla had a product that could compete with Microsoft. The question now was how to publicize it. Open-source management seemed to work well when it came to developing software. But open-source marketing was an unexplored concept. Would people be willing to volunteer to get the word out on Firefox?
Baker thought they might, and one reason was Asa Dotzler. Back in the late 1990s, Dotzler had been a young college dropout working in a market research firm in Austin when he joined Mozilla's community of "bug reporters"--people who volunteered to download the latest prototype of the product and submit technical reports of what needed fixing. Dotzler, who does not have a technical background, took it upon himself to act as a liaison between nontechies willing to serve as bug reporters and the programmers--helping to get the former up to speed and the latter to be more forgiving, and thus bringing thousands of new people into the project. Baker was so impressed that in 2000 she offered him a job. "I said, 'You want to pay me for something I'm spending 30 hours a week on as a volunteer? Are you kidding?"
Now Dotzler wanted to see if the nontechnical community would be willing to help promote Firefox. The company created a Spread Firefox website, and Dotzler helped make it the focal point of a launch campaign. The goal seemed absurdly ambitious: Generate a million downloads of Firefox within 10 days of its release. Dotzler and a crew of Mozilla faithful asked users to identify blogs that might be receptive to the initiative. Soon, thousands of blogs and other webpages were sporting "Download Firefox" buttons.
Released in November of 2004, Firefox Version 1.0 hit the million-download mark in four days and the 10 million mark in 30 days. A few thousand volunteer marketers had joined the campaign. Two months later, about 30,000 people were helping. Then a volunteer made a suggestion: What if the community could get, say, 10,000 people to each throw in $30 to buy a New York Times ad for Firefox that would list the names of all the contributors? Within 10 days of opening up the offer, the list was close to 10,000, and in December the Times carried a two-page ad. Over the next year, the number of marketing volunteers climbed to more than 100,000 and Firefox was being downloaded at an average rate of 250,000 times per day.
Mozilla's headquarters are next to Google's (NASDAQ:GOOG) in Mountain View, California. But where the latter company's imposing office complexes twist through acre after acre like a 21st-century industrial version of a topiary maze, Mozilla's digs are buried in a corner of a small village of modest wood-sided structures vaguely reminiscent of ski lodges or barracks. The parking lot is full of aging Honda Civics and Toyota Corollas. At the front office, there's no receptionist. A piece of paper suggests you come on up and knock.
Computer workstations abound, but so do toys and stuffed animals, and a cozy canteen offers a neatly arranged cornucopia of berries, nuts, and cereals, as well as an electric massage chair and a cappuccino machine. There's a lot of energy coursing through these offices, and a chunk of it comes from the nine student interns currently in residence. One of them is John Carey, who was a Missouri State University student when Mozilla plucked him out of the Midwest and set him up with a free apartment nearby, along with the assignment to make a video of the behind-the-scenes work to prepare the next major version of Firefox. He readily agreed, but also professed an interest in arranging for a crop circle in the shape of the Firefox logo. "We thought he was kidding," says Dotzler. In August, the 40,000-square-foot logo was pressed into an oat field near Salem, Oregon, with the help of a crew of college students, making it one of the largest, not to mention more offbeat, advertisements ever created.
The offices are also home to Mozilla's management. But many of the company's 70 full-time employees are based in Toronto, Tokyo, and Paris. It's not unusual for a firm to have to wrestle with problems presented by a geographically scattered work force, but in Mozilla's case the challenge is extreme. That's because what drives Mozilla isn't the 70 employees but the nearly 200,000 volunteers who do most of the product development and marketing.
Open-source software organizations such as those that developed the Linux PC operating system and the Apache Web server program have long wrestled with ways to keep lines of communication open and to distribute decision making among vast communities of mostly software-developer volunteers. But Mozilla can't fall back on these precedents, either. Unlike other open-source ventures, which tend to be niche products embraced by techies who become fiercely loyal to and dependent on the software, Firefox is a mass-market, consumer-oriented product that can easily be replaced should it fail to offer distinct advantages over the competition. That means Mozilla has to move faster and be more innovative and marketing-oriented than its open-source cousins.
But the way Mozilla really strains the conventional open-source model is that it makes money--pretty good money. Baker won't reveal the details, but the company is said to have pulled in revenue of around $70 million last year, most of it from a deal with Google in which the search engine giant kicks back a commission on ad revenue obtained when a Firefox user enlists the browser's Google search bar and ends up clicking on a sponsored link. Mozilla has struck a similar deal with Yahoo (NASDAQ:YHOO). In spite of some grumbling about how money would "contaminate" Mozilla's noble mission, the Firefox community has seemed to approve of the opportunity to bring in revenue. But the community was not intimately involved in the decision-making process. How could it have been? Negotiating a big deal with a large commercial partner requires speed, decisiveness, and, most ticklishly, confidentiality. Managers at high-tech giants these days won't even discuss sports until a nondisclosure agreement has been signed.
And then there's the question of how a not-for-profit foundation like Mozilla can even consider these kinds of commercial deals, given the higher public disclosure requirements for nonprofits, not to mention the restrictions on money-making activities that accompany the favorable tax status. To get around these barriers, the Mozilla Foundation in August of 2005 spun off most of its Firefox-related operations, including Baker and most of her management team, into a wholly owned for-profit subsidiary, Mozilla Corp. As a private, for-profit firm, Mozilla Corp. can cut deals and keep secrets. But because it's owned by the foundation, any profits go to furthering the foundation's mission of preserving choice and innovation on the Internet.
Plenty of companies, of course, can and do claim to be driven by mission. Google still beats its chest over its "Don't be evil" credo, and Microsoft will talk endlessly about its dedication to the user experience. But executives at these and all other publicly held companies are obligated to use whatever legal means they can to generate the most profit possible; if they don't, shareholders will replace them with managers who will. Mozilla Corp. managers, in theory, don't face these pressures.
That's one reason they can continue to distribute decision making to the community, open-source style. Indeed, Baker seems convinced not only that Firefox devotees deserve to be involved in everything Mozilla does, but that they will do a better job than can a small group of clever managers. "I have a very optimistic view of the social fabric," she says.
This, of course, poses some tricky management questions--issues that go beyond the obvious ones of long time frames and a lack of confidentiality. Baker points out two such problems: "stop energy" and "loners." Stop energy, as she describes it, is a sort of gravity that permeates group decision making, the inertia that sends good ideas crashing down to earth, stifling innovation in the process. Loners, by contrast, are like unguided missiles--they simply take it on themselves to push ahead, without regard for what others think. Loners may get a lot done, but unchecked, they can take a toll on the community by creating resentment and divisiveness that ultimately slow progress.
To counter these problems, Baker has championed an approach she calls "layers of the onion." New decisions start off in small, relatively specialized groups that spring up informally within the community as needed. These groups can filter out weaker ideas and improve promising ones and gather information that can inform decision making. Since the groups are self-selecting and likely to contain highly motivated members, they're resistant to stop energy or loner behavior. The best ideas from these smaller groups are then passed on to a somewhat larger group that can further refine them. Only then are the most promising ideas and most important questions put before the entire community.
The result is a hybrid organization, one that relies heavily on open-source models of distributed decision making and transparency but that is also, when necessary, capable of keeping secrets and of making quick and potentially controversial decisions.
Baker acknowledges that the model is far from perfect. "Mozilla is inventing a new sort of entity, one that's different from organizations we know," she says. At one point in 2005, criticism from the community that Mozilla managers were being too secretive led Baker to hold a moratorium on all corporate-only meetings for several months; it only ended when managers needed to discuss a human resources question too personal to share with everyone. "One of the biggest management challenges is to not let this core of Mozilla employees take over," Baker says. In other words, Baker is struggling to limit the influence of her own management team--not the typical goal of corporate executives.
So far, however, the model seems to be working. One way to gauge its success has been Mozilla's continuing ability to draw more and more marketing volunteers. A contest to come up with promotional videos drew 300 entries, some of which have garnered millions of views on YouTube. The Firefox logo has been etched into the lawns in college quads, spray-painted on highway overpasses, baked into pancakes, sent up 100,000 feet on a weather balloon, and crayoned onto countless human faces. Seventy thousand websites have put up a "Download Firefox" button. Some 180,000 people have offered to help develop, test, or promote the software. Dotzler notes that up to 10,000 volunteers help spot bugs, and he hopes to build a network of volunteers willing to provide technical support via telephone, IM, and e-mail.
Firefox now commands about 15 percent market share for browsers; it has been downloaded more than 200 million times. But it's facing a renewed challenge from Microsoft, which has brought out Internet Explorer 7 to match many of Firefox's features. Mozilla managers like to suggest that Firefox will continue to thrive because users still distrust Microsoft because of its long neglect of the browser and the resulting security fiasco. Ross Rubin, director of industry analysis at high-tech market research firm NPD Group, questions that assumption. "Most consumers aren't religious," he says. "They'll use whatever tools are most convenient for them and that work the best for them." But he adds that IE7 doesn't seem to introduce any significant new features that would cause Firefox users to jump ship. "If Firefox is serving their needs, they probably wouldn't switch back," he says. Still, few observers think Mozilla can continue to win converts at the astounding rate of the past two years.
But for Baker and much of the Mozilla team, it has always been about more than building a better browser. It's the future of work and competition that's at issue. More leaders, Baker contends, are recognizing the ways in which communities can drive organizations. The question is not whether this approach can work, but how best to implement it. "I think the answers will apply outside the software industry," she says. And many observers agree. "Companies tend to treat consumers as having to be manipulated to make money off them," says the University of Washington's Krishnamurthy. "Mozilla's message is that if we all work on this together, and we allow people to express their frustration with the status quo creatively, we can take back control. Any underdog company can feed into that."
Harvard's O'Mahony believes that some companies are already borrowing from Mozilla's playbook by trying to get consumers not merely to swallow marketing pitches and buy products but also to help design products and shape and disseminate pitches. Topps' (NASDAQ:TOPP) Bazooka bubble gum brand, sneaker maker Converse, and MasterCard (NYSE:MA) all have asked the public to come up with videos and other marketing-related efforts; Procter & Gamble (NYSE:PG) and Kraft Foods (NYSE:KFT) have looked to consumers for help with product development. "Mozilla has shown that organizational boundaries can be more permeable and transparent and that you can develop a relationship with a community that takes an interest in how you organize and manage," says O'Mahony. By the same token, she adds, Mozilla also is blazing a trail for not-for-profit organizations that could benefit from behaving more like for-profit businesses.
O'Mahony notes that many entrepreneurs might have trouble following Baker's lead when it comes to openness. "She listens, she debates, she changes her mind, she adapts, she allows herself to be convinced by outsiders," she says. Don't take O'Mahony's word for it. Anyone can have a window into Baker's inner debates via her expansive blog, which is unusually thoughtful and candid. In one entry from last year, Baker mused on how an entrepreneur typically has a broad vision--she calls it a "sense of possibilities"--for a company that prompts a leader to try to exert control over everything that goes on. "My case is very much the opposite," she writes. However, she adds, "there is of course a piece of the Mozilla project where I have a very strong sense of the possibilities and a determination to see things proceed in a way that makes sense to me. That area is the organizational structure of the project. How do we integrate the various constituencies? How do we organize ourselves? How do we provide enough structure to build top quality products and still provide room for individual initiative and serendipity?"
If fretting over the quest to find the ideal organizational structure is Baker's favorite blog topic, her second favorite would be her long-standing effort to master the trapeze. In one entry, she describes trying to nail a maneuver 22 feet up, repeatedly failing until she finally identified the kernel of fear within her that was keeping her from kicking hard into the flip. She faced down the fear and did it right. At such moments, she writes, "…a moment of 'float-i-ness' or seeming lack of gravity appears. At these times it feels like one has all the time in the world. Last night I learned again how a small amount of fear, seemingly too small to matter much, has far greater impact than one might imagine." Here, and elsewhere, Baker seems to be hinting at the source of the resistance to the effort to do something that is fundamentally different than what has been done before. What is stop energy, after all, but a fear of flying?
David H. Freedman is Inc.'s technology columnist.