Many pundits are railing about Cisco's move into the server market because the company's current gross margins are 10-20% less in the server market than in their core router market. What these critics miss is that the evolution of computing is to create corporate clouds of functionality and storage – and those who control the architecture of the cloud will dominate the commercial relationship. (This post is also at the HBS site.)
What the heck does it mean to have a "cloud"? It means that the functions you want like storage, communications, and applications, don't need to live on the device you have in front of you. You can get them on demand from the network. This is an old idea and aspiration, but it is starting to come to fruition for four fundamental reasons: first, there is much more demand for portability, and due to power and weight constraints, the device you carry with you will continue to ask for functionality from the network, as the iPhone does, not have it local to you as your PC does. More lightweight PC tablets and smart phones are sold than PCs and the difference is increasing. Second, the bandwidth, at long last, is starting to be there so that a connected device's connection is not so slow that it kills the experience. 3G cell phones are barely fast enough to have a decent network experience. The new developments in bandwidth promise to create a much better network experience because the response time will be there.
Third, at any given time it is estimated that 85% of the computing capacity of the work is sitting idle, consuming hundreds of billions of dollars in power. There is no doubt that it is much more efficient to run all your computing as one big network, that can load balance the capacity – rather than having lots of fragmented machines, so there are significant cost savings and economies of scale to be had by going to a cloud approach. Fourth, and most interestingly, the IT industry now puts out as much carbon as the airline industry, and IT is growing at a much faster pace. If we have a carbon cap and trade system as the Obama administration desires, there will be significant incentives to go toward cloud computing because as Larry Smarr, director of the CalIT lab puts it – we will need greener bits. Amazon, Google, Microsoft, and Yahoo! All use cloud computing and have data centers in states with high percentages of hydro-electric power so that they can have greener bits for their customers. It is for these four reasons I believe that the promise of cloud computing will finally be here.
Well, if I'm right about the above, then Cisco, like any self respecting strategically aggressive, innovative, market leading company, wants to try to control its own destiny by capturing what Kim Clark many years ago called the "strategic control point." In any technology industry, there is a set of intellectual capital and organizational capability which gives you the high ground from which you can extract higher economic rents. Microsoft did it in operating systems. Oracle started with high performance databases, and added ERP. Cisco is realizing that as HP and IBM begin to provide integrated cloud computing, that those large firms will try to marginalize the role of routers to be simply a part of their overall solution – eventually eviscerating Cisco's margins. Cisco, rightly is holding onto its strategic position in their account relationships.
Is this move risky? Absolutely. Do they have a choice? Sure, but they will have a lot less bright future if they are willing to cede the architectural control of the customer and their offering. Also, I'm sure they have a few tricks up their sleeves like the ability to provide superior bandwidth and more storage across the network than the competition – which means that it might be IBM and HP that feel the pain, not Cisco.