There is an article over at Infoworld about virtualisation ROI that is as interesting for the comments it has attracted as the piece itself.

The writer, Tom Kaneshige, makes the point that creating persuasive value metrics for the value of server virtualisation can be tricky, especially for small IT shops. One respondent takes issue with the numbers cited. It's interesting, provocative stuff.

Seasoned ROI experts will know, of course, that anybody can make numbers sing, dance and maybe even do somersaults if there is sufficient motive. There's nothing new there: it has always been the case that there are lies, damned lies and statistics, and that many people use statistics the way a drunk uses a lamp post - more for support than illumination. Vendors who supply ROI calculators make sure that everything will look rosy; FDs who don't want to spend will find any jot of data to support their stinginess. The truth will lie somewhere between these polar positions.

Virtualisation is particularly tricky. The promise of future savings is immense and many companies see what they regard as a saving inside 12 months. On the other hand, virtualisation licences aren't cheap and there can be significant associated costs.

Smart CIOs will do what they can to measure investments and returns in the full knowledge that they are not dealing in empirical fact. Incidentally, CIO columnist Andy Hayler recently wrote a fine piece, here, on the subject.

However, case studies and networking with peers will be highly valuable in ensuring that virtualisation rollouts follow best-practice guidelines.