From my desk at Xconomy San Francisco—aka my Potrero Hill apartment—I can look out the window at Highway 280, watch the traffic that crawls into the city every morning and back out every evening, and chuckle smugly. There’s never any traffic congestion on the commute from my bed to my coffee machine to my laptop. I’m a remote worker, and I love it.
Given that Xconomy is a distributed news network with editors in eight cities, almost everyone in the company works this way, except for the folks in our Cambridge, MA, headquarters. It’s a successful model for us. Not having offices keeps our overhead low, and ensures that our staffers spend a lot of their time out in the community, where they should be.
But here in San Francisco there’s a much more famous example of a distributed Internet company where remote work is the norm. It’s Automattic, the company behind the WordPress open-source content management system and the WordPress.com blogging platform. Automattic has an “office,” but it’s really more like a lounge, where local employees can come and go as they please and the company can hold occasional group events.
WordPress began in 2003 as a long-distance collaboration between a pair of programmers named Matt Mullenweg and Mike Little, and the pattern continued after Mullenweg started a company to expand the platform in 2005. By the end of 2006, Automattic had assembled a team of 18, drawn mostly from the community of volunteer WordPress contributors. But they were scattered around the world, and Mullenweg had not met many of them in person.
Even after Automattic instituted annual meetings where everyone got to meet face-to-face, most employees continued to work independently. And that’s how the company was still organized when Mullenweg invited Scott Berkun to join Automattic in August 2010.
Berkun is a Seattle-based project manager, consultant, and author who cut his management teeth leading parts of the Internet Explorer team at Microsoft in the 1990s. He’d done some consulting work for Automattic, and he feared that the company had gone overboard in its devotion to its engineers’ autonomy. Building great new features for WordPress.com would be easier, Berkun told Mullenweg, if the company organized engineers and designers into small teams. The team members could still work remotely, but they ought to coordinate their projects and report regularly to leaders, Berkun recommended.
Mullenweg agreed to try out teams as an experiment—and asked Berkun to lead the first one himself. Berkun agreed, on the condition that he be allowed to write a book about the experience after his tour of duty. He’s done that, which is how the world now has a book called The Year Without Pants: WordPress.com and the Future of Work. It’s an essential read for anyone who’s wondering if their own organization might benefit from a dose of decentralization; tempted to try being a remote worker themselves; or just curious about where WordPress.com came from and how Automattic really works.
A student of organizations, Berkun writes that he arrived at Automattic feeling like “an old dog in a futuristic workplace.” He wasn’t sure whether he’d be productive as a manager working remotely from his team members, or even whether the traditionally flat culture at Automattic—with everyone reporting to Mullenweg—was ready for the idea of teams.
By the time Berkun left in January 2012, his team of four had proved its value by introducing some key new social features into WordPress.com. They’d gathered physically a few times—notably in Athens, Greece, for a few days in November, 2010—but otherwise they’d worked from their respective homes in Seattle, Los Angeles, San Francisco, and Ireland, coordinating their task lists and schedules via a combination of Skype, IRC (Internet Relay Chat, a 1980s-era chat protocol), and blogs. (Automattic employees engage in a constant online conversation about company projects using a collection of custom internal blogs known as P2s.)
Berkun concluded that distributed teams work just fine—as long as the people on them are disciplined, self-sufficient, and so passionate about their projects that they’d probably want to work on them even if they weren’t getting paid. That’s a pretty high bar, obviously; the supply of such people is limited. So it’s not clear to me, just judging from Berkun’s experiences inside Automattic, that giving employees extreme levels of autonomy would work well for most organizations.
On the other hand, it works great here at Xconomy, and I’m completely on board with what Berkun calls Automattic’s “results-first” culture. “It didn’t matter if you were pantless in your living room or bathing in the sun, swinging in a hammock with a martini in your hand,” he writes. “What mattered was your output.” If more companies respected their workers enough to measure them this way, we’d probably have a far happier workforce and a faster-growing economy.
I interviewed Berkun this week, and asked him to talk more about how The Year Without Pants came about, what he learned at Automattic, how technology undergirds remote work, and what he hopes readers will take away from the book. What follows is a slightly edited version of our conversation.
Xperience: Is it true that you knew going into your year at Automattic that you wanted to write a book? If so, why did you want to write about them specifically?
Scott Berkun: Yes, that was the plan all along. I’ve known Matt on and off for years. A couple of years ago, he had me do some consulting for Automattic. He invited me to some company meetings and asked me to give him some feedback about how things were going and how I thought they could make the company run better. So I got a little bit of a sneak peek at the company, and one thing I thought would be interesting to do was to organize into teams. It was not rocket science to make that suggestion, but it was one of the pieces of advice I gave them to help improve productivity and address some of their other concerns.
A few months later, they decided to go head and do that. They contacted me and said ‘Hey, we are going to form teams, and we need some people to lead them. Do you want to come in and try out this whole idea of yours and manage one of the teams yourself?’ That is the high-level story.
The personal story is, I have written four books and have been doing this author-speaking career thing for about a decade, and have been successful doing it, but it’s a career laden with hubris. It’s easy to lose sight of how the working world is changing. Many people who write books and give lectures don’t ever go back into the fray and see how much of their own advice they can actually practice. I was thinking I should probably go back and manage a team again at some point. I didn’t know when or where, but when Matt suggested this idea, I said this could be it.
Then, thinking through all the crazy, unusual things they do at Automattic, it seemed like a fantastic way to test my own ability as a manager in an environment that was so novel, with no e-mail, an open vacation policy, open source origins—there are all of these things that many people think are the future of work. So it seemed like a great project to kill two birds with one stone, and get my next book out.
X: Going in, were you skeptical about the feasibility of remote work and all this decentralization?
SB: Yeah, I went in skeptical about many things, and I am still a little skeptical in some ways. I wasn’t sure that it actually works, or that it’s a thing that should be emulated. And I wasn’t sure that the notion of forming teams would work in an organizational culture that was so passionately anti-corporate, anti-structure, anti-hierarchy.
X: So, what did you learn? Did you come away less skeptical?
SB: I think one of the biggest things I learned, if I had to pick just one, was about hiring. For all the criticism about how the company actually functions or whether [distributed work] is right or wrong, they do such a good job at hiring that it compensates for most of these other issues. They fact that they hire by trial [all prospective Automattic employees must complete a project before being brought on full-time] is a wake-up call for most of the work world, about how broken our interviewing and hiring practices are.
X: What about remote work itself—what were your before-and-after attitudes about how well it works at Automattic?
SB: I didn’t really have a strong opinion going in. I was skeptical about it, but I think what I learned was that it works fine, if you have a smart team of talented people who are motivated to be collaborators. When it doesn’t work I don’t think it’s because of the tools; it’s because the people aren’t talented, or they don’t like working with each other, or they are given bad direction or leadership by whoever is managing them.
My year there totally proved that remote work can work fine for teams. We did all kinds of different projects, some of which required schedules with dependencies, and some of which were more ad hoc and agile. But if it’s a good team that’s motivated to work and you have the authority to manage them well, then they can do a good job.
X: For companies thinking about whether to implement a more distributed structure and allow remote work, what would you say are some of the biggest benefits and pitfalls?
SB: If the goal you have as a leader is to have a culture of people who are passionate and independent and autonomous and aggressive about pursuing ideas, distributed workforces make that likely, because these are people who have to manage their own time and work independently. If, on the other hand, you have a 1970s, MBA mindset and you want command and control and you want to be the center point of all decisions, then distributed work is probably a mistake, because by its own nature it diffuses people and their attention and your decision-making authority.
X: It seems like it’s the kind of structure that would only work at companies where the founders or top executives are comfortable being hands-off, the way Matt Mullenweg and Automattic CEO Toni Schneider seem to be.
SB: Yes, but it’s a philosophy for an executive that I would encourage at any kind of company that’s ambitious about progress. I have experienced managers who are like that—very hands-off—even at Microsoft. Gabe Newell, the CEO of Valve, has a similar philosophy, and Valve is not a distributed company—they do many other things that are unusual but one thing that’s normal is they all work in the same building. So there are definitely others with a similar “benevolent dictator” mindset.
X: You point out in the book that new management fads come and go, and that it can be disastrous to try to graft a structure like remote work onto an organization that doesn’t have the right culture for it. I interpreted that as a kind of “don’t try this at home” warning. But in other places you seem very admiring. Which side do you ultimately come down on? Is Automattic-style decentralization something more companies should emulate, or are you saying it’s probably a bad fit for most organizations?
SB: Your question is a good one, and I totally understand it, but I wouldn’t say either is true. I think the book should be an eye-opener and a call for managers to experiment. There is very little risk in letting an employee say, ‘You know what, I want to try working from home for the next month.’ A manager could say, ‘Okay, why don’t you try that, but I am going to evaluate you on these specific things, and if you can meet those performance objectives, then great, go for it.’ That kind of experimentational attitude is what every manager should have.
The book isn’t meant to be a bible on how to be Automattic. It should be a bible for managers who want to try new things. That is the only way they are going to achieve the ambitious goals they have—by being open to trying new ideas. So I don’t have a hard line about recommending that every company should switch to being distributed. But I do have a hard line about this: if you are managing things, what new things are you trying? Automattic offers a whole suite of things to use. If Automattic did these 12 interesting experiments, many managers and executives should be able to try at least one or two. That’s what I hope people get from the book.
X: Lots of individuals are probably attracted to the idea of working remotely from their organization. What kinds of people are cut out to succeed at this, and what kinds are not?
SB: There’s a chapter in the book that tries to talk about how this all fits into the future of work. In there, I offer that that the people who do better at distributed companies or remote work are the ones who have a little bit more self-discipline. They don’t require being in the same room with a bunch of other people who are working to be motivated to work. They are good at managing their time. There is just more independence in an environment like this, and less social structure to depend on to get work done. So people who struggle to want to go to work and manage their time and manage their attention will probably have a harder time, because they don’t have a boss checking them every day and looking over their shoulder, or a coworker reminding them to do this or that. They are more autonomous.
X: That must be part of the point of Automattic’s hire-by-trial process—finding the people who can work autonomously.
SB: Yes, it fits with that. The project that you get, you have to manage yourself. There is no one there giving you a step-by-step on what to do every day. You are expected to go and do it on your own. A lot of talented people are uncomfortable with that much autonomy and that’s an indication that you probably wouldn’t do well in an open, chaotic environment like Automattic’s.
X: One of the sub-themes in the book is about the blogging and chat tools, like Skype and IRC and the company’s internal P2 blogs, that Automattic employees use internally to stay connected. You write near the end of the book: “It might be that progressive companies like Automattic are open to the idea that technology can return some of the meaning of work we’ve lost.” Can you explain a little bit more what you meant by that?
SB: It’s liberating when none of these tools are official or mandated—when employees are told, ‘We hired you and you know what you’ve got to get done and it’s up to you to decide what tools are going to be most useful.’ That gives meaning to people who are passionate about their work—that management and the company itself is not a hindrance, not a source of friction. Instead it’s a support structure.
Automattic does a great job at this. For all of my criticism about P2 and IRC—and a lot of the people in the company agree with me and use different things at different times—there are no draconian rules about using these tools. You’re free to choose. Which is no surprise, given the open-source roots of WordPress itself.
X: Speaking personally, e-mail is the scourge of my existence—there’s just far too much incoming stuff, and dealing with all of it takes too much time away from my actual work. On the other hand, maybe that is my actual work. And it’s not clear that any of these other tools would work in our organization—we’ve tried some and they haven’t caught on internally. It sounds like you’re not a huge partisan of any particular technology as “the answer.”
SB: I’m really ambivalent about every tool that exists. They can be abused, and they can be used well. The culture around them and the habits of the people you work with really define how well they work.
In most organizations, the experience of e-mail is poor because e-mail is abused. Too much information is sent over e-mail, when there are better alternatives for certain kinds of information. But it’s hard to blame e-mail technology itself, which hasn’t changed much over the last couple of decades. We have changed, and our habits have changed.
It’s really hard to change a culture that is centered on one tool to another tool. It requires leadership and clear notions about what the problems are that motivate the changes. You’d have to highlight for people the frustrations of e-mail and argue that these are the kinds of communication that would be better held over here [using another tool]. Blogs and comment boards are a much better way to deal with certain kinds of communication. But all tools have value, and all tools can be problematic depending on how they are used.
By posting a comment, you agree to our terms and conditions.