Gartner has a press release out on its latest UC findings, and there's some interesting stuff there. I have to say, I think they really get it in terms of identifying the key challenges and pitfalls as we move into the critical time for UC adoption.
Gartner has a press release out on its latest UC findings, and there's some interesting stuff there. I have to say, I think they really get it in terms of identifying the key challenges and pitfalls as we move into the critical time for UC adoption.Specifically, Gartner seems to have done some significant examination of cost factors for all the elements that will support a UC deployment--not just the servers or applications, but application development and bandwidth:
Gartner predicts organisations that don't follow best practices for involving the networking team with IT projects will pay at least twice as much for their application development projects and will still deliver poorly performing applications. "Ideally, networking should be involved from the beginning of the application project, actively proposing functions that can be usefully performed in the network, as well as ensuring that applications are tested," said [Gartner VP Steve] Blood.
Gartner also predicts that between 2007 and 2011, organisations will overestimate their applications' bandwidth needs and waste £21.6 billion on gigabit Ethernet and a further $2 billion on powering it.
Gartner also predicts that between 2007 and 2011, organisations will overestimate their applications' bandwidth needs and waste £21.6 billion on gigabit Ethernet and a further $2 billion on powering it.
A couple of points there: Note that Steve Blood stresses the importance of the applications team involving the network people in app development projects. Networking people have been rightly encouraged to learn more about SOA, Web services and other issues related to application-enablement, but as Steve Blood points out, the process has to be two-way. And as Tom Nolle has written here before, the applications people are to a large extent driving the IT train, so it is somewhat incumbent on them to get the networking folks aboard.
Another point in that excerpt is the notion of overpaying for bandwidth. It's an interesting point not exclusive to UC. The notion of "throwing bandwidth" at networking problems relied on the premise that LAN bandwidth was essentially free, that once a 10/100/1000 NIC and server module reached cost parity with the old 10/100, you might as well go for the extra bandwidth. But with energy costs on the (most likley) permanent upswing, lighting up that extra bandwidth may not necessarily be "free" any more.
Gartner's final recommendation likewise makes a lot of sense:
Mr. Blood recommends that organisations looking to implement UC should create a virtual team of all the stakeholders, both from the business units as well as in IT and look at which groups of users can benefit most from improved collaboration. He added: "Those that have already adopted UC should now verify that productivity benefits are being realised and that users are taking advantage of the new capabilities. Deploying UC is expensive so companies should ensure those that have it are using it."
