I’m at PC Forum and wanted to jot a couple of ideas that I heard here that I thought were worth writing down.
John Seely Brown and John Hagel led a moderately interesting discussion of loose coupling vs tight integration. Their main example was Chinese motorcycle companies, which were competing with Honda for the world motorcycle market. By loosely coupling with their suppliers, and working collaboratively, they were able to quickly innnovate the price of a motorcycle down from $700 to $200, while Honda was still at $900. There were two major reasons it worked: loosely coupled modules in a business process can be rearranged or reintegrated to form new things or try new ideas much more quickly than tight integrations. Also, innovation within a module is possible without disrupting the entire system. One other caveat: this mechanism needs a reference model to work from, in this case a Honda motorcycle. Other cases cited include Linux (working from the Unix reference) and Apache (from NCSA, among others).
One interesting company was SoloMio. They provide services through cell operators to allow quick response to an incoming call without necessarily answering it. Imagine when a call comes in from your wife, you can quickly select an option that responds with “I’m in a meeting – can I call you back in an hour?” This is a great idea, and it’s too bad it requires so much work on the part of the operators to implement. One idea I want to add to this is one that’s been kicking around in my head since about 1999, for a Bluetooth watch. No, I don’t want to get the time via Bluetooth. I want to know who is calling (caller ID) and have a set of 3-5 responses I can send with one touch. That way, when someone calls, I can just glance at my watch, and possibly press a button, to respond – with minimal interruption of the conversation or meeting I’m currently in.
There was also an interesting open source panel, with Kim Polese (CEO of SpikeSource) and Mitchell Baker (Chief Lizard Wranglers of the Mozilla Foundation, and boy does she look different from her picture in the program). They both represent large Open Source projects with corporate sponsorship. One thing we didn’t explore in too much depth was the shift (or maturity) from random developers from scratching their own itches and gratifying some ego, to real enterprise application development. To my mind, there is not always much of an intersection between the people with domain knowledge in an enterprise business process, and people with development capability. So, what is interesting is that in the new models it’s companies that are altruistic, not just people. Some companies have the farsight to realize that it is in their long term best interest to pay for the development of software that will be given away.