Life and death of a company based on Open Source

This was probably the longest interview in the book and interesting because ArsDigita was the first company in the Founders series that established a successful business around distributing a free, Open Source product. At JumpBox, we like Open Source - every application we distribute is Open Source and we’re debating Open Sourcing the JumpBox technology itself. So Phillip Greenspun’s story hits home for us. He founded the company in 1997, grew it to revenues in excess of $20MM in five years and then had an unfortunate snafu when they chose to take $38MM in venture capital and were ousted from the board. They did an end-run around the VC’s and recaptured control of their company but the damage had been done and they eventually chose to sell their stake. The company folded shortly thereafter in 2002. This story is an interesting one because most of the others it ultimately failed. We can typically learn more from mistakes than successes however, so what we can we learn from Phillip’s roller coaster experience?

How they got started

ArsDigita emerged from a photo blog project that Greenspun started to document his travels on a road trip to Alaska. He started sending email updates and eventually published his updates as web pages with photos from his trip. He started getting decent readership and most of the questions centered around photography, so he answered them and published his answers but this correspondence generated twice as many new questions. He eventually setup a discussion forum and found that other readers would chime in and answer questions for him. A digital photography community site was born.

He had written custom software to run his community site and decided to post the code powered the site as an open source project. He began receiving calls from big companies that wanted to use the software but required customizations. Greenspun was still at MIT working towards his PhD but after receiving a string of calls with big companies offering him $100k contracts for two weeks worth of work, he decided that his paltry salary as a grad student and the rule about “no dogs at work” was enough and he put school on hold to pursue the consulting opportunity.

Their secret sauce

They started as five programmers working out of a rental house and each was essentially doing his own project working independently but sharing the improvements to the codebase. It was a loose federation of consultants and what differentiated them was their emphasis on the notion that a programmer was a professional and could carve out a name and reputation for himself. They pushed profit/loss responsibility to the individual programmers for ensuring that their project was completed to the customer’s satisfaction and then they paid massive bonuses to the ones that produced happy customers. There were no bloated teams with middle management and dark recesses of an organization in which people could hide if a project slipped. All projects had tiny teams of no more than three people per job which lead to clear accountability and transparency. They also had a strong mentorship and apprentice program - they believed in developing a programmer as an agent would develop an athlete.

Greenspun was also very much aware of the advantage of being a “craftsman vs. a factory worker” and the work alienation and quality degradation that occurs when the programmer interaction with the end customer is removed. In speaking of traditional programming environments, Greenspun says, “The programmers were in a corner doing what they were told. That’s one reason they were so easy to outsource. If a programmer really never talks to the customer, never thinks, just solves little puzzles, well, that’s a perfect candidate for something to offshore. So I said, ‘I don’t want my students to end up like this. I want them to be able to sit at the table with decision-makers and be real engineers… to be an equal partner in the design, not just a coder.”

We have a similar philosophy at JumpBox in that everyone answers support requests. No engineers reside in ivory towers- everybody touches the customers and gets the praise or criticism first hand so the experience is direct and tangible. There’s no game of telephone with sales people and customer support representatives to mangle the message from the customer and relay it to a programmer.

Their guerrilla marketing tactics

ArsDigita never ran ads in major publications nor did they engage in traditional PR (while Greenspun ran it). They promoted almost exclusively on an educational platform of tutorials on their web site, books in bookstores and seminars. Greenspun says, “Almost all of our marketing and sales was educational. We just thought, ‘We’ll teach people stuff, and some tiny fraction of those people will become our customers.’ It seemed to work just as well as running ads, which were a hard sell and kind of empty and a waste of people’s time. In this case, nobody could ever say that we wasted their time. I think the same percentage of people that read an ad in ComputerWorld magazine and bought something would read one of our tutorials and buy something from us.”

We’re about to engage in a similar tactic with JumpBox. The strategy will become apparent when we release the individual Open Source application tutorials soon but it’s the same notion of teaching people something useful but piggy-backing our JumpBox product on the lesson to show how it becomes even easier. Our thinking is the same that the tutorials themselves are valuable and will be transmissible for their standalone utility but they will also become an vehicle that carries awareness of our product to new potential customers. Most importantly, this tactic helps us to reach the “people that didn’t know what they didn’t know” before finding out about JumpBox. Making that leap out of the domain of people who already understand the benefits of virtual appliances is critical and is the hardest chasm to cross. But the people you reach when you do so are the most excited by what you have to offer.

They were their own biggest customer

The entire time they were customizing the ArsDigita community software for their clients, they were incorporating the changes into their own instance that was running Photo.net. So for instance if a change they had made had unforeseen performance implications that only surfaced under heavy load from users, they would know before the client’s site went down because it would manifest on Photo.net. Greenspun says, “By contrast, companies like Microsoft were still developing software for the Web as if the Web didn’t exist.”

The “eat your own dog food” approach is a recurring recommendation in startup stories. We intend to cut our public site over to be run on a Wordpress JumpBox as soon as we bring our stuff online in the new data center (right now it runs on a traditional Wordpress installation). We envision offering an entire “JumpBox Office Suite” of applications at some point and of course we’ll string together the ones we use ourselves and be our own first client. For now we’re laser focused on the march towards sales. The RC3 release we did this week brings us into the home stretch and the next thing we produce will be a sellable product.

The mistakes that led to the decline

ArsDigita was a private, profitable with $20MM in yearly revenues and $3MM in profits. They were making so much they had to look for ways to spend the cash to lower their tax basis so they would do things like pre-pay a year’s worth of rent. Unfortunately employees would read news stories of other programmers that had started some random dotcom and made millions before the bubble burst. Greenspun felt the pressure to IPO the company and when he researched the process he found that the underwriters wouldn’t even talk with him because their incentive is tied directly to the fees they collect on doing IPO’s and the time required to conduct due diligence on a company that had NOT taken venture capital made it comparatively unattractive when they could just go do deals with VC-funded companies and skip the due dilligence. Greenspun was essentially pressured into taking VC in order to IPO the company to satisfy the people that were reading all the rags-to-riches stories of entrepreneurs making it big. Greenspun was a CTO at heart who had found himself in the CEO role so in one sense he welcomed the idea of offloading this responsibility to someone who was (theoretically) a natural CEO.

The two VC firms (Greylock and General Atlantic) installed rookie members on their board, guys with MBA’s who had never actually run a company. Greenspun says, “The guys on my Board had been employees all of their lives. You can’t turn an employee into a businessman. The employee only cares about making his boss happy. The customer might be unhappy and the shareholders are taking a beating, but if the boss is happy, the employee gets a raise. By contrast, the businessman cares about getting a customer, taking his money, not spending too much serving that customer, and then selling something more to the same customer. These are totally different psychologies.”

There’s about fifteen pages of the interview dedicated to telling the story of the debacle that occurred with the VC’s, how they ousted Greenspun from the company and how Greenspun eventually snatched back control via a loophole in the agreement. The Founders book is worth reading for this story alone. Rather than recount the long, bloody tale of what went down, let’s focus on the essence of what went wrong: the puppet CEO extricated the programmers from the front lines and made the “factory floor” error of distancing the craftspeople from their end customer. Beyond all the errors of hubris, management and accounting mistakes, the core failing was in restructuring the company culture on the axis of how programmers interacted with clients and had accountability for the profit & loss of their individual projects.

The biggest lesson

“The one thing [I would have done differently in retrospect] would probably have been slightly slower growth, I guess. Not to worry so much about the competition, concentrate on getting really good people who shared the company’s vision, who could be mentored to the point where they could then recruit somebody else. Basically, just limiting growth.”

Greenspun sold his remaining shares of the company in July of 2001. ArsDigita closed its doors in January of 2002. Greenspun got his pilot’s license and now focuses on his passions of photography and being a helicopter instructor.

This entry was posted by Sean Tierney on Thursday, May 24th, 2007 at 2:53 pm and is filed under Blog. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.