Innovation is risky. It may cost more than it is worth.
Innovators take a high risk for a potentially high gain. Early Adopters take a lower risk for a lower gain. Mainstream takes minimum risk for the gain of keeping abreast of the rest. Those behind the mainstream take a higher risk, that of losing out to mainstream competitors. Sometimes the gain of not needing to spend to keep up with all mainstream advances can justify this risk.
Any method of increasing the gain or likelihood of gain, and reducing the risk, in one’s chosen approach to innovation, is to be welcomed. Necessarily, flexible and responsive approaches are better for dealing with any risk. Here we deal specifically with those approaches that relate directly to Innovation. The latest technologies are always held to have value in their own right; conversion to OO, or to components, to web technology, EJB or to patterns, is considered "good" without a second thought. Even keeping in the mainstream requires adopting innovations as the mainstream adopts them, and this bears its risks, as does all innovation.
Clearly the first thing to do is decide where one is comfortable with respect to one’s approach to Innovation. Are we Mainstream? Early Adopters? If trailing the mainstream, is this deliberate policy for a known purpose, or is it stagnation and decay?
When the organization is leading the innovation, and to a lesser extent when introducing innovations as Early Adopter, Mainstream or whatever, consider reducing the risk by isolating, testing and evaluating the innovation with a SkunkWorks team.
Modern approaches avoid unnecessary innovation by saying YouAren’tGoingToNeedIt.