Is Comcast going to buy Yahoo?

A good deal has been written about Brian Roberts, CEO Comcast, announcement that 160mbps connections are just around the corner. The problem is the internet was built on best effort relationships and is nothing more than a bunch of networks connected together with no quality control on where or how these various networks interconnect. Without some standard for interconnections it is impossible for Comcast to guarantee any performance of packets once they leave the Comcast network. The network is as strong as it's weakest link so even though you may have the ability to get 160mbps, chances are you won't if the origin of your requested content is not physically in/on your AS.

IMHO this is more about setting the stage for a new approach to the comcast network than bandwidth to the internet in general. It is in Comcast's best interest to get as much capacity on their own network as possible because it gives them a huge advantage over the likes of Google, MSFT, YHOO, Fox, etc when it comes to QoS on specific applications. Besides, there is no possible way this could scale out to the internet and that is on purpose....meaning comcast will never be able to guarantee 160mbps anywhere but on their own network. These ever evolving broadband local networks with some overlay backbone connecting them together is the ideal situation for p2p distribution. At the end of the day, comcast wants to keep those packets on it's network as long as possible because then and only then do they maintain 100% control. In doing so they have leverage with the customers who are mainly home users. Service for the enterprise is not far behind nor hard to accomplish...a simple dedicated line between the enterprise and comcast's network and the enterprise is connected. Or better yet, the enterprise could outsource their IT services such as email, CRM, storage, collaboration, voice, video conf. etc. to Comcast and be guaranteed XX Mbps throughput from an employees homes to the corporate environment. Given all of the local and federal incentives for going green which will trickle down to driving companies to promote telecommuting even more than they are now, this is actually a valuable service to be able to offer. Add video conferencing, voip, VOD, etc on top of it and they'll have a network that offers an enterprise customer the most efficient way to address all of their voice/IP/IT requirements. In doing so it also de-stresses comcast's reliance on upstream links by using them less(assuming p2p is in place) and relying more and more on peering. Comcast was one of the early peering sluts so to speak, when other MSOs thought peering was what you did when you teamed up with your second grade classmate to do a bookreport, Comcast was signing network peering deals with anyone who sends a fair amount of traffic to them or vice versa.

I have mentioned Surewest previously and they are currently offering fiber to the home in Sacramento. For customers where this is available, everything (voice, HDTV, IP, VOD, catv, etc) are delivered over IP. It may seem cutting edge and far too complicated for a dumb old cable company to accomplish but comcast is no TCI even if they are saddled with some of the TCI infrastructure, Brian Roberts gets it and embraced it long ago. If that isn't enough, IP really is proving the most efficient delivery medium available. @home was developing products and applications around these virtues and Comcast has merely expanded on that vision.

IMHO, so take it for what its worth, Comcast and Google are the two powerhouses today. Comcast should buy Yahoo or better yet, pull a trifecta Comcast/MSFT/YHOO combo. That would be the google killer. Why bring in MSFT? They have lots of cash, a great revenue stream, lifelong customers and can turn office apps into a hosted infrastructure which yhoo no comcast have done. All of these services may even be free to the end user because they'll all be subsidized by advertising. After all, it is advertising that drives goog's revs and it will be advertising or subscription fees(I view them as the same as they should be mutually exclusive) that surpass licensing revenue of software vendors. Remember, advertisers are fickle and will spend their dollars where they think they are going to get the best return. As such, when a comcast user needs to access google servers, it is comcast who controls how well that user gets that packet back and it isn't necessarily in their best interest to make it smooth and seemless. Think about it, causing poor user experiences with google is one way to discourage ad spend with google. One piece at a time and sooner or later it all adds up to a significant threat to google's so called "unfair advantage"