Open source may be a venture capital dry hole

The clock is ticking toward midnight. A lot of players were funded in 2007. Three years in the financiers are getting impatient.

I'm afraid this is a feature, not a bug. As Dave himself notes, most start-ups from the class of 2007 are still too small to be worth a purchase by a big dog.

The reason should be obvious by now. Most of the savings from open source go to customers, not financiers. You give away the software and, in hard times, most downloaders find it more expedient to make things work themselves than to pay for support. Or they rely on the community.

This can be a false economy. Even inside programmers cost money. What's wrong with hiring some time from the experts? But most support contracts can't commit the time of committers to a specific customer. There are only so many hours in a man month.

Still, some companies manage to make money. Red Hat makes money.

Most people credit Red Hat's success to its relationship with IBM and the fact it's selling a Linux. But I suspect there is more to it than that.

Marc Fleury built JBOSS around committers, and many cashed out when Red Hat came in during 2006. It has taken Red Hat a long time to build back a proper support infrastructure, but now that work seems to be finally done and, while the company doesn't break out its numbers, JBOSS seems to be a profit contributor rather than a cash sink.

It may be that the process Red Hat undertook these last few years is the real "secret sauce" of open source success.

The best open source companies are service businesses, like law firms. Just as restaurants sell food for more than they pay for it, so open source companies sell the time of their people for more than they pay for it.

This doesn't mean, as some think, that you just give customers access to cheap people. It means you build a structure that will make mid-level people appear to be as fine as committers. This can be done. A good quality assurance staff can be worth its weight in gold, and a good manager can monetize that asset.

Of course this is an East Coast business secret. It's not how Silicon Valley rolls. It takes time, training, and organization to build this kind of infrastructure around a software product. Maybe Sand Hill Road doesn't have the patience for that nonsense.

Too bad.

Maybe more open source companies should consider moving east?