Marketshare Redux

Back in May, when I was still at Ingres, I wrote a blog entry on the measurement of market share in the software industry. One of the comments I made welcomed Gartner's new way of looking at market share that gave a little more credit to the open-source companies who don't charge license fees.

Another article has come out recently about how IDC is predicting that the Linux ecosystem will reach $40 billion by 2010. I picked up on this one by reading Matt Asay's blog entry on the subject.

Now, I have a love-hate relationship with the analysts. Many of them I consider to be close friends and I've told them that I think analysts are usually predicting the past and half the time they get that wrong.

The major issue I have with the IDC numbers is they're using old techniques to measure and judge the success of an open-source product/project. The concept of open-source projects is their ability to remove money from the ecosystem, not in creating a new, large(r) ecosystem. I realize that investors want to know how much money they can make by investing in open-source companies, but if you really want to be useful then tell me something I don't know.

I know that Linux and other open-source businesses are going to grow tremendously. I know that you can look at dollar figures over a time-line and predict future dollar figures.

What I don't know is how much money is being removed from the IT infrastructure budgets. Open-source is about doing it better, cheaper, faster and more transparently than the closed-source vendors can allow. Tell me how much money will be saved by the IT industry by adopting open-source vs. closed-source and you'll give me a much better picture of the impact of open-source.

The closed-source vendors have trained the analysts to measure success by how much money they make. How about measuring success by how much money customers save? Now, that's a prediction I'd like to see someone make.