Virtualization would be easier to explain if it was just about consolidating hardware. Thankfully for your business, it's not. Virtualization's ultimate benefit is not consolidation but IT agility and speed. To get there, you'll need to do some painful things. HP's John Bennett recently pointed out to me one truth that many IT leaders miss in the current marketplace noise of the VMware v. Microsoft fight: If you don't pay attention to your applications portfolio, your virtualization effort will hit a wall.

It's not enough to trim down your number of physical servers in the data center, says Bennett, HP's worldwide director, Data Transformation Solutions. You must trim the fat from your application mix as well.

Oh, I know, the danger lights are flashing for many of you. Touch the line of business applications? That's a dangerous and complex project, one that many IT leaders will turn away from for political reasons. Simply consolidating the current apps on fewer physical servers seems much safer.

Make this choice and you're cheating yourself and the business, Bennett says.

"Consider what you could do if you did an application modernization project before you did your virtualization project," Bennett says. "That's when you get to multiply results instead of just adding them."

"It's very tough of course," Bennett adds. Indeed. Plenty of IT leaders understand the need to get rid of old applications that require dedicated staff expertise, and often, specialized boxes with low utilization rates.

But application portfolio trimming gets messy and complicated fast. Maybe that's one reason why only a handful of the CIOs I've talked to about virtualization in the past year actually did application portfolio trimming as part of a virtualization effort. (The other reason: many of you got an adrenaline rush from the consolidation benefits of the first wave of virtualization, and completed that wave quickly to show a win to the business. With that win under your belt, maybe you now have more political capital to tackle those ugly old applications.)

Forward-looking CIOs know that the application mix must be refined for efficiency's sake. Who wants to be balancing workloads and carrying costs for applications that are barely used or should have been put out to pasture years ago?