Report from “Making a Business on Free Software”

(This is cross-posted to the OSDDDI blog)

I attended the Berkeley Entrepreneurs Forum's (BEF) session entitled "Open Source: Making a Business on Free Software" last night (Note: this is not the Stallman-sense of "free software").

The panel included chair Philippe Cases (General Partner of Partech International), Zack Urlocker (Marketing VP for MySQL), Jacob Taylor (CTO and Co-Founder SugarCRM), Paul Doscher (President & CEO Jaspersoft) and Allan Leinwand (Operating Partner at JP Morgan Partners).

I'm going to paraphrase the extensive notes I took in this post. It was a pretty fascinating event; both in substance and in a cultural-sense. (Since this post is rather long, I'm going to include the rest below the fold.)

[More:]

First the cultural stuff. I don't go to many business school events and last night I felt like I was sticking out like a sore thumb; I was wearing a t-shirt and the rest of the room was filled with about 200 button-shirted, dockers-wearing MBAs, professors and Haas alums. Another strange ritual was that, apparently, before each BEF, they "run the numbers". I figured this would be tracking the performance of former entrepreneur's work or new firms. However, it appeared that this was a moderated opportunity for business people to pitch 90 seconds of material to the crowd. About seven open source start-ups pitched their product and asked for help with business models, MBA interns and even coders (the only one I remember was koders).

Now, on to substance. The questions from the chair centered around three themes:

• What is truly innovative about OSS? How do we compete?
• How do we build community?
• What is different about sales and marketing of OSS?

What is truly innovative about OSS?

As to the first question, Mr. Urlocker from MySQL answered that it's the instantly global reach that an OSS product achieves. MySQL doesn't have to hire developers in the Bay Area; in fact, they've got developers in 21 countries and they don't have to pay most of them US wages. The way they do their hiring is to look at people already contributing quality stuff and ask them if they'd though about doing it for a living. MySQL has about 200 employees and there is already about 200 books out there on MySQL. In short, it's a lot more than code contributions, it's about community and empowering people.

Mr. Taylor pointed out that within 24 hours of the 1.0 release of SugarCRM, some developers had already translated it into french. They had thought it would take a few months to get that exposure and hadn't yet implemented language flexibility capability (language packaging). Mr. Taylor then went on to say that the community is "very happy to take things the rest of the way" and improve their product for them in terms of code contributions, interoperability with other products and just general feedback.

Mr. Doscher said it was simply that OSS allows them to be quicker, cheaper and faster than other firms doing the same thing with proprietary software. He said if you put the kind of control behind an organization that exists with the ability to modify code they run to run their business, you end up with a lot more power as a OSS vendor than you'd have just inside the walls of your own business.

Mr. Leinwand said the key was flexibility. Because you have users you've never met or even had a business relationship, your reputation is your code. Business are starting to realize that they've been oversold consistently on the quality guarantees of opaque software. They want to see (or have someone look at) the nuts and bolts of a system before they pay for it or make a decision to make it part of their business.

How do we build community?

Mr. Taylor said that it was hard to release the code the first time. However, they hadn't seen the combination of skills and experience in other open-source CRM products that they felt they could bring to bare. It took about two months for SugarCRM to go from zero points on sourceforge to the top ten list and then they were project of the month. He said he felt that interoperability and facilitating interoperability is key as it works to bring people from all types of other competing and complimentary software.

Mr. Leinwand had talked to another data networking client who had spend $80,000 for a piece of gear and then separately stitched together 4 open source projects on a stock PC that did that provided the same functionality and performance for $300. However, when asked this client thought it would die if they released their integrated OSS solution publicly. Mr. Leinwand realized that they needed some pieces of the commercial business infrastructure to to sustain it and to catalyze the community.

Mr. Urlocker said that the important question was how to get momentum. As opposed to having to beg for usage, MySQL gets people using it without them asking. And then they come and pay for it. He said, if you can get people that consult, document, etc. your software, you create a lot of people that have a vested interest in your success. There is a lot of excitement when people start to transform there lives; a guy who's aspiring to become an Oracle DBA has an uphill battle as he has to have a lot of capital to even get started. As said before, people are fed up with being oversold with crappy software. With good software, the viral effect will do the rest. He finished by saying that we are now seeing a level of professionalism in OSS that exceeds the quality of what you can get from proprietary software. That sells well.

Mr. Doscher said that as an open source project or product, you need an enemy; but you never want to mention who that enemy is. This, believe or not, helps to build a community. You have to have someone to shoot for, to rally against. If you say it publicly, you'll get crushed; if you frame it as community (us vs them) you solidify the community.

What is different about sales and marketing of OSS?

Mr. Doscher began by saying that it's all about transparency. Your main communications vehicle is your website. On your website, you'd better have a lot of _useful_information like demos, a product roadmap, etc. Developers are smart people and they need to be able to play with things to be convinced. He said if you try and take a different approach, it will fail. Developers want all the information there and they're pretty good at convincing their bosses that something works or doesn't work; when they really need your help, they'll call.

Mr. Taylor acknowledged that SugarCRM's biggest hiccup was that they weren't communicating properly at first. He said, strangely enough, making sure that people understand your business model is key. People will freak out if they don't know where you're going, how your going to get there and that you won't die on the way.

Mr. Urlocker said that they really don't have much marketing other than their brand. They let the brand and install base do it's work and then, we people start asking, they jump in with their salesforce and marketing team. He said, people pay you when they're in pain. He repeated a maxim from an old job: "The only thing worse than someone pirating your software, is someone pirating someone else's software."

Joe's Take

I was really surprised at how little these guys mentioned licensing and licenses. In fact, I asked a question:

One thing that you guys haven't mentioned at all is licensing. What kind of advice can you give us about choosing a licensing model? Specifically, before any code has been written or if we find ourselves with a patchwork of source from different projects?

I was really surprised that they misinterpreted my question to be about per-seat licensing. Two of them said, "We operate on a per-server licensing model." Mr. Urlocker actually understood the gist of my question and said that this needs to be very carefully considered and that it pays to think about it early and with very good legal representation.

All in all, I was surprised that 200 MBAs, etc. were sitting, on the edges of their seats, to hear about these open source businesses. Some of it seems to be with the new entrepreneurial rush on OSS. For example, the Chair's firm, Partech, didn't invest in any OSS business until about one year ago. He realizes that this was a big mistake and that they should have been in it 2-3 years ago in order to be better positioned for the current climate. The take away for the MBAs seemed to be: Think twice before starting a purely proprietary software company.