I really wish Dave would write more often. Yesterday he wrote about analysts and the value they provide, as well as their track record on open source. The gist? Analysts get paid to interpret the past, not the future, and so have missed the significance of open source:
I truly believe that it is exceedingly difficult to use one's analytical skills and past experiences to predict new waves and new technologies. New things come from out-of-the-box thinking; they don't come from tried-and-true methods.Dave is right. Back when I was at Lineo (embedded Linux vendor), I remember how frustrated I got with the analysts covering the space, because every prospect to whom I talked was rolling out Linux somewhere, on some device, or at least had it in pilot. But you saw almost none of this data in the analyst reports. To them it was Symbian, Windows, VxWorks, etc. There was no concept of the deluge that was about to happen.It's a huge stretch to assume that a mainstream analyst who has spent their career analyzing proprietary software companies, and understanding how those companies value their closed-source technologies as intellectual property, is going to easily transform into someone who understands the benefits of open-source.
These analysts think that software companies only exist because they have proprietary, closed-source, secret sauce ingredients that would be too expensive for anyone else to copy. Therefore, an open-source company can't possibly succeed as a software company.
In similar fashion, I can tell you from direct, personal experience that open source applications are being rolled out all over the planet in copious amounts, by decidedly "late adopter"-type companies. Financial Services firms, yes, but also in Manufacturing, Media, Government, etc. etc.
Yet most analyst firms are still preaching last millenium's message.
Of course, Dave may be right: it might be our fault for improper nomenclature:
Perhaps the real issue we have is that we should stop calling open-source companies software companies. Maybe it's our own fault for using obsolete terms to try and get others to understand what we're all about.The problem is that the Purchasing department continues to think in familiar terms of "per server," "software," and such. But this can change over time. And, oddly enough, it might be the SaaS companies more than the open source companies who succeed best in changing the way we define ourselves.
Posted by Matt Asay on April 27, 2007 06:01 AM
