At the beginning of OSCON in his keynote address, Tim mentioned a difference between software and services that caught my attention: if you buy a piece of software and the company goes out of business, the software still works. On the other hand, if you take the people out of a business like Google, or even your favorite ISP, there's no more service---its just goes away. This isn't a huge revelation, but its an interesting way to think about the service economy being about people rather than things. I had this in the back of my mind as I was listening to Doc yesterday.
Doc was talking about the construction business. He points out that we use construction industry metaphors all the time when we talk about building computer systems. That's an interesting perspective and meshes with Tim's comments. Doc talked about driving through some industrial area somewhere and noticing business after business with huge lots full of pipes, structural steel, and the like. These things are commodities and businesses that sell them make good money (albeit not with the kind of margins that Microsoft and Oracle have promised their shareholders). Moreover, the construction industry is large, profitable, and honorable business. Doc thinks this is a model for where the software industry is headed. I agree.
The construction industry is about service. While we typically don't thing of construction being part of the service economy, I think that view concentrates too much on the things and not enough on the construction itself.
I built a house several years ago and my general contractor definitely spent his time providing a service. Sure, he built things too, but mostly he assembled commodity products to build a custom house for me and that didn't diminish his ability to create value and be compensated for it. The companies who supplied the commodity products made money too.
Tim talked about a paradigm shift in his keynote. As I listened to the talks at OSCON, this thought kept coming back to me over and over again: building service-based businesses on commodity software products isn't just an idea for a business model, its the primary business model of computers in the networked era. Yahoo!, Google, ISPs, and other successful net-businesses are using this model right now and doing quite well at it. What's more, open source software is getting more and more capable all the time. Combine these facts with Tony Perkin's belief that now is the cheapest time ever to start a net-based business and I think you're staring opportunity in the face.