Virtual Broadband ISPs.  The Internet service provider (ISP) business for AOL, EarthLink, and MSN is changing.  It's in a transition from a bundle of network connectivity and software to a software only business.  The old model consisted of:

1. A free ride over the last mile.  In the old model, the ISPs could ride on the subsidized last mile dial-up connections of the Tel/Cable cos. 
   
2. Private networks.  ISPs ran their own modem farms, LANs, and connections to the Internet (and their central server farm).  
   
3. Centralized services.  Since Internet connectivity cost money, it behooved the ISPs to run private networks that connected to a central hub of services.  The more time customers spent on the private network, the lower the costs of outbound Internet connections. 

The new broadband model has put this traditional approach into a tail-spin.  The Tel/Cable cos and cable companies now tightly control the network connection from the last mile to the Internet.  The new environment is characterized by:

1. The rates the Tel/Cable cos. are charging for connectivity is expensive.  Broadband rates range from ~$40 to $60 a month with virtually all of that going to the Tel/Cable cos.  This leaves little wiggle room for ISPs to charge more for services.
   
2. Private networks and centralized services are a liability.  Since Internet connectivity is bundled in the basic package offered by the Tel/Cable cos., there is little incentive to shunt users via a private networks to centralized services.  These costs now represent a huge burden to the ISPs.  It gets even worse as the ISPs attempt to add rich media content to their service offerings to attract people to broadband.
   
3. ISPs don't have to pay for network.  The flip side of not being able to charge for the network means that the ISPs don't have to pay for the bandwidth used by their customers on Tel/Cable cos. unlimited usage plans. 

The ISPs are now forced into a new mode of operation which is similar (but in many ways worse due to the parallel infrastructure costs needed to maintain dial-up customers) to that experienced by the Internet portals during the Internet boom.  To avoid the inevitable bloodbath that this situation will result in, ISPs are going to have to change their strategy.  Here's how new "broadband" ISPs will change the game:

1. Phase out infrastructure to radically lower costs by moving to the desktop.  A heavy desktop software package that does much more of the work on the desktop will enable the ISPs to take advantage of the ~130 m PCs that were sold this year with GHz processors and copious Gb hard drives.  Work done on the desktop means fewer servers at the central hub and less use of their private networks.   The faster these centralized systems can be downsized (and head count reduced), the more profitable ISPs will be. 
   
2. Focus on connecting desktop users to improve services at no additional cost.  ISPs should focus on riding on the "free" Internet bandwidth provided by the Tel/Cable cos to connect customers across networks.  Desktop publishing tools, P2P connections, and rich desktop clients can be combined in a witch's brew to provide compelling shared media experiences.  This improvement can be done at little to no marginal cost. 
   
3. Market the hell out of it.  With a lower cost structure in place, these new "broadband" ISPs will be able to radically improve margins while adding services.  Higher margins will support higher per customer acquisition fees, which will allow "broadband" ISPs to outspend rivals on marketing and promotion.  

12:05:20 PM