This is a question I hear quite often: how did you manage to bring Funambol to profitability, push the open source project to 500,000 downloads and get your company included in the Red Herring 100? I have no clue, really, so my standard answer is: "I am a lucky guy"...
Since my answer usually ignites another question ("no, really...") and I usually look stupid, I decided to put together a few ideas on what I would do if I had to start an open source company from scratch. Here they are.
1. Think BIG: open source is a game of volumes. You need a BIG community for all the positive effects to show (quality, support, contribution, viral marketing, quick sales cycles, international reach, ...). The only way to get a big community is to attack a large market. If you build an open source project for a niche, your community will never get really big and your company will always stay small. I know I am killing half of the open source companies that VCs have funded in the last year or so, but ...
2. Think STANDARDS: standards have been key for a lot of successful open source projects. HTTP for Apache, SQL for MySQL, J2EE for JBoss, SyncML for Funambol (not in that league yet, but we are working on it...). I am not saying that being the open source implementation of an open standard is the only way to create a big open source project (I feel SugarCRM might make it without it), but it certainly helps. A lot. In particular, if the standard is successful.
3. Think COMMODITY: this goes with the BIG argument above. Open source is the winning strategy for commodity markets. It is a game of volumes. Price is key in commodity markets. If you choose something that will become a commodity, at the end you will come on top because nobody can compete with open source companies on price. Not because it is free (it is not), but because our companies are built on different cost structures than proprietary companies. Our marketing, sales, business development and QA costs are a fraction of those of proprietary companies. In a commodity market, it really matters.
4. Think PRICE CUSHION: if you want to create a big open source project, you do not need it. If you want to create a big open source company, you need a price cushion. You have to go after some big proprietary company which set a very high price. Being open source, the price for your product will always be perceived as a little bit more than zero. How much over zero makes all the difference. If you have someone big that set a high price, you'll crash that price slowly until the market becomes a commodity and volumes are there. If you are thinking about attacking a large market where there is no price cushion and no big player charging a ton, you are going to create a big open source project but zero dollars for your shareholders. Think MySQL with Oracle, JBoss with BEA, SugarCRM with Salesforce.com, Funambol with RIM (that's $50/user/month...). Once again, I know I am killing the other half of the open source companies that VCs have funded in the last year or so, but ...
5. Think IRONMAN: most people that went public with a proprietary company will tell you "it was a marathon". Open source companies are in the Ironman category. That is "SWIM 2.4 MILES! BIKE 112 MILES! RUN 26.2 MILES". The race ends with a marathon, but you have to swim a lot and bike a ton before you even start the marathon. Aggregating a large community takes years, there is nothing money can do to accelerate the process. It is a natural long process with phenomenal fruits at the end. But you have to be patient. MySQL was first released internally on May 23, 1995... That's eleven years ago and they might go public this year or next (go Marten!). Do not think starting an open source company will be a quick race, gear up for the long long haul. The Funambol project was started in 2001 as Sync4j, and we reached 500,000 downloads in 2006. The Ironman cap was the first company gift we ever produced. One of our key employee is an Ironman competitor (just won the 12 hours of Cesate, go Daniele!). When people ask me where do you think your company is in its life, I answer "we just got out of the water after the swim fraction, we are biking downhill, but I am bit worried about the marathon that we'll have to start in a few years".
I hope I finally answered the question: it is really tough and I AM a lucky guy.
