Jason Fried & 37 Signals

The Brash Boys at 37signals Will Tell You: Keep it Simple, Stupid
By Andrew Park, Wired Magazine

To the 300 software developers packed into a Vancouver conference room, David Heinemeier Hansson was more than a programmer. He was a visionary, the creator of Ruby on Rails, a software template that powered an increasing number of hot Internet applications. He was a philosopher-king whose minimalist ethos suggested a new way of thinking about business and software. And he was a celebrity, with boyish good looks, precocious self-possession, and fans who invoked his name so frequently they used his initials as shorthand: DHH. As Hansson took the stage at the British Columbia Institute of Technology for this, the first Ruby on Rails conference, the room was filled with the kind of giddy excitement that greets the opening chords of a Hannah Montana concert.

The program billed Hansson's keynote as a collection of "beloved rants" and "favorite tales from the land of righteous indignation," and he didn't disappoint. He began by congratulating the nascent Ruby on Rails community (and, by extension, himself), citing a litany of impressive achievements: 500,000 downloads of the code, 16 how-to books, mentions in Wired and other publications, and several industry awards — including, for Hansson, the prestigious Hacker of the Year title, bestowed by Google and O'Reilly Media.

But not everyone was convinced of Rails' revolutionary potential. Critics had been saying that Rails wasn't versatile enough, that it couldn't handle large amounts of traffic, and that Hansson himself was arrogant. "Arrogant is usually something you hurl at somebody as an insult," Hansson said. "But when I actually looked it up — having an aggravated sense of one's own importance or abilities' — I thought, sure."

Then he clicked over to the next slide, white letters against a dark background that spelled out his response to the naysayers: fuck you. The crowd erupted into laughter and applause.

Hansson's programmer-with-a-messiah-complex shtick may be a hoary cliché. But in the nearly two years since he delivered this presentation, he and his partners at software developer 37signals have backed up the big talk. Rails has continued its run of popularity; over the years, tens of thousands of programmers have used it to create countless online applications, including podcasting service Odeo and microblogging phenomenon Twitter. And Basecamp, 37signals' Rails-powered, easy-to-use online collaboration software, boasts more than 2 million account holders. Signal vs. Noise, the 37signals blog, pulls in 75,000 readers a day. Hansson and 37signals cofounder Jason Fried are "revered," says business author Seth Godin. "They are as close as we get to demigods online."

What's more, the pair's once-heretical vision — that there is beauty and wisdom in Web-hosted, bite-size software built to accomplish narrow tasks — has become conventional wisdom. In the two years since Hansson's keynote, Google released Apps, the relatively feature-free alternative to Microsoft's bulky Office suite; Facebook opened its platform to independent developers, unleashing a stream of mini-applications that offer everything from playlist-swapping to Boggle bouts; Salesforce.com's AppExchange gave corporate software developers a platform for selling tiny, downloadable programs; widget wunderkinds like Slide's Max Levchin and RockYou's Lance Tokuda became Web celebrities; and venture capitalists opened their wallets in the hunt for the next little thing. "Simplicity is the most important thing in technology," says Paul Graham, cofounder of early-stage venture firm Y Combinator. "And it's only getting more important."

None of this has helped Hansson discover any hidden wellsprings of modesty. He has called Microsoft "entirely optional," referred to Java as "grossly overused," and described Flash applications as "horrid."

But if Hansson hasn't changed much, neither has the programming framework he created or the business he heads. For some, that's a problem. Hansson and Fried have steadfastly refused to grow their company, beef up their products, or explain their plans for the future. Now, critics argue, the pair's reactionary embrace of all things minimal has made their products less useful and could cost them influence, customers, and millions of dollars.

Hansson has a predictable response to such charges. "I don't usually go around saying 'Fuck you' to everyone I meet," he says. "But sometimes it's the appropriate answer."

The defining characteristic of Ruby on Rails is, as the name suggests, speed. Using Rails, an adept programmer can create a simple blogging application in 15 minutes or a photo database in five. Two guys built Twitter in two weeks.

In exchange for that speed, programmers accept a Hansson-knows-best approach to software design. While most programming languages require coders to build every new application from scratch, Rails gives developers a set of configurations that lets them bypass the busywork. That makes Rails ideal for quickly creating lean, sparsely designed Web-based applications, which coincidentally enough is exactly what Fried and Hansson think software should look like: as Fried puts it, "stripped down to the absolute bare necessities."

Fried developed his theory of streamlined software design in 1994 as a junior at the University of Arizona. He was looking for a simple database program to catalog his music collection. "I downloaded a bunch and they all sucked," Fried says. Instead of focusing on the relatively easy task, they were overloaded with options that only complicated the process. "I said, I can make this way better.'" So he created his own program, dubbed Audiofile, and peddled it as $20-a-pop shareware, earning enough to keep himself in beer money.

After college, Fried returned to his native Chicago, where he formed 37signals — a Web design firm, named in esoteric reference to SETI — and posted a manifesto on his homepage that railed against the shortcomings of most software. ("The Web should empower, not frustrate," he wrote. "Just because you can doesn't mean you should.") On his protoblog, Signal vs. Noise, he further developed his philosophy. "Remember — size does matter: A small group of 10 great people will outproduce, outwork, outthink a large group of 50 average people."

Fried's missives struck a throbbing nerve, and before long Signal vs. Noise was drawing a dedicated readership of programmers and designers similarly fed up with bulky, inelegant code and enthralled by Fried's edicts. It was through his blog that Fried met Hansson: In 2002 Hansson, then a student at Copenhagen Business School, provided some programming advice after Fried posted a question about the best way to handle pagination using a programming language called PHP. The two became fast friends. "Our outlook was the same," Fried says: "Keep it simple." So when Fried wanted an online collaboration tool for his employees, he again turned to Hansson. Working 10 hours a week over four months, Hansson wrote the code to support Fried's spare, airy interface. Hansson used a little-known language named Ruby — which most developers felt was too slow and limited to be of much use — and developed a series of shortcuts to help him build the program quickly and easily.

The result was Basecamp, a lean but effective platform requiring no costly servers, tricky installations, or technical support. Although he'd developed it for in-house use, Fried realized the commercial potential of the program after showing it to friends and clients who wanted an inexpensive and simple way for small teams to work together. When he released Basecamp in February 2004, Fried expected the monthly subscription fees, which today range from $12 to $149, to generate sales of $5,000 a month by the end of Basecamp's first year; they reached that target in six weeks. Five months later, Hansson packaged his Ruby shortcuts and released them as Ruby on Rails, which started winning converts almost immediately.

At the same time their software was taking off, so was the duo's cult of personality. In 2005, Fried gave a 10-minute presentation at Tim O'Reilly's Web 2.0 Summit, the influential confab of some of the Internet's biggest minds. The blogosphere lit up
with praise. (The response was so overwhelming that Fried himself posted a blog entry wondering if 37signals had "jumped the shark." Commenters leapt to his defense.) In 2006, the company compiled a list of contrarian dictates — don't plan, don't hire, don't fix every bug — and published it as Getting Real, to rave reviews.
But the key to Fried and Hansson's burgeoning celebrity may have been their $895-a-seat workshops at which acolytes celebrated the gospel of radical simplicity. After attending one, Ryan Norbauer was inspired to tear down Lovetastic.com, a successful personals site that he had spent eight months creating in PHP, and rewrite the entire thing using Rails. Now Norbauer runs a Rails consultancy. "Rails has become a very big part of my life," he says. "I don't think I would be doing programming for a living without it."

That kind of devotion is common. After Sean Tierney read Getting Real, he bought 10 copies for his employees at Grid7, an application development shop, and insisted they read it. "Jason Fried is a genius," says Tierney, who today runs a software startup called Jumpbox. "He's the opposite of everything corporate."

Tucked away on a grubby side street in a gentrifying loft-and-warehouse neighborhood about a mile west of downtown Chicago, 37signals' offices hew to the company's small-is-beautiful edict. Actually, offices is a strong word: Headquarters consists of four desks shoved up against a wall. 37signals leases its 500 square feet of floor space from a design firm whose employees surround 37signals's work area. There is no 37signals sign, no receptionist, no indication that 37signals even exists. The company has just 10 employees, five of whom telecommute and none of whom are expected to work more than 40 hours a week. But 37signals hasn't remained small out of sloth or through lack of opportunity; indeed, it's taken some effort to keep it from growing. Fried says he has rebuffed numerous inquiries from venture capitalists looking to invest in his company. (The sole exception: Amazon.com founder Jeff Bezos, whose investment firm, Bezos Expeditions, took a minority stake in 37signals in 2006 for an undisclosed amount. The company has said it accepted the deal because it offered access to Bezos, not because of the money.) Neither will Fried entertain acquisition offers. "Someone on the outside would look at what we do and say, Let's ratchet it up to some enterprise level,'" he argues. "I don't want to make our software more complicated. I really don't understand why everyone's interested in Fortune 500 customers. I just don't get that."

37signals may not be targeting corporate behemoths, but its pared-down offerings may be inadequate for even its smaller clients, some of whom have urged Hansson to adapt Rails so it is better suited to handle popular applications. In March 2007, a Twitter engineer told an interviewer that he was having difficulty getting Rails to handle his company's massive spike in traffic. Hansson responded by sending a heated email to Jack Dorsey, Twitter's CEO, and chastising the company on his blog for playing the "blame game" instead of solving its scaling problems itself. (The two firms have since resolved the dispute.) In January, an executive from hosting provider Dreamhost mused about the difficulty some of his clients were having running Rails applications. Again, Hansson responded on his blog: "Wipe the wah-wah tears off your chin and retract the threats of imminent calamity if we don't drop everything we're doing to pursue your needs."

This sort of hostility can't come as a surprise to anyone who has followed Hansson or Fried, but there are signs that their churlishness is beginning to generate some backlash. The Basecamp message boards are filled with complaints from unhappy users, fed up with the software's paucity of features — functionality of the Opera browser, say, or better version tracking of uploaded files — who have switched to competing products. "They take the position that they're right and everyone else is wrong," says Douglas Karr, director of technology for an Internet marketing firm, who stopped using Basecamp in April. "It really just put me off the company." Harper Reed, CTO of online T-shirt retailer Threadless, says that the belligerence of Rails' followers soured him as well. "It's very much like a religion," he says.

What's more, 37signals' ideological objections to outside funding could make them less able to withstand competition. Nicholas Carr, author of The Big Switch, says companies like 37signals won't have the resources to fight should larger firms with huge economies of scale and backend infrastructure decide to take them on. "They're going to have a very tough challenge," he says.

Fried says he doesn't worry about losing individual Basecamp customers, since none of them pay more than $149 a month. He points out that the company's total revenue doubled in 2007. And in addition to Basecamp, 37signals' other products — subscription-based programs like group-chat app Campfire, content management tool Highrise, and information manager Backpack — pull in hundreds of thousands more users.

But, faced with a seemingly endless buffet of appetizer-size software, industry insiders have begun to question the basic philosophy that Web-based mini-applications are inherently better than their bulkier but more powerful competitors. "Running your application on Rails places a huge limit on what you can do," says Charles Forman, founder of iminlikewithyou.com, who has abandoned the framework for Merb, a rival programming tool. That promises more scalability. A recent survey by the NPD Group found that fewer than 1 percent of desktop PC users had replaced a desktop application — such as Microsoft Office — with a streamlined online alternative like Google Docs, even though the latter is free. Design expert Don Norman, a consultant for Microsoft, says that one reason for the disparity is that customers actually like and use the extra features. "Complexity is a necessary byproduct of the modern age," he says. "When you actually sit down and analyze what you need to get the job done, it's not simplicity."

That's heresy to Fried, Hansson, and their followers. Call it arrogance or idealism, but they would rather fail than adapt. "I'm not designing software for other people," Hansson says. "I'm designing it for me."

Andrew Park (andrewpark4@gmail.com) is a business writer in Chapel Hill, North Carolina.