The Hype Curve measures public perception of a given technology. The points on the curve are labelled in terms of perception and opinion ("buzz of expectation", "slough of despond", "renaissance of hope" and "liberal enlightment" - or something like that anyway).
Surely the sensible advice to CIOs is to discount public perception, and base technology judgements as far as possible on technological reality rather than hype.
- Don't adopt something simply because it's popular.
- Don't NOT adopt something simply because it's NOT popular.
It's certainly possible to find cautionary Gartner statements to this effect. But the emphasis of Gartner advice (at least as glossed by V3) seems to contract the second point. Don't adopt Twitter, says Gartner, because the hype curve will turn against you. In other words, base your technology judgement on the hype curve after all!
Apart from CIOs who are so stressed-out or weak-minded that they like safety in numbers, the people who really value this kind of prediction are the software vendors. For them, the hype curve (as well as the Magic Sorting Hat) may help predict their likely product sales as well as the effectiveness of a given marketing campaign, and help plan their investment in certain products.
And of course Gartner wants the hype curve to be perceived as accurate and relevant. Gartner benefits commercially if the majority of CIOs gregariously (herd-like) follow the Gartner advice. "Thinking with the majority". Even if it is confusing hype with reality.
Update: See also Geek and Poke's Gartner Hype Cycle Version 2.0 (via Cloud Ave)