I was struck by a comment to my last posting, in which I asked \u00a0for reasons why CIOs aren\u2019t part of the innovation process. It got me thinking about whether IT departments are trying to innovate in a vaccuum.Here\u2019s the key portion: \u201cMost IT people are trained to shoot holes in user requirements (ideas), in order to refine them and make them feasible in a computing framework. We are trained to automate what was already being done by the business\u2026 Innovation requires that you understand the business and its strategies-- and that you look for ways to change the playing field in some way. We as IT professionals must leave our area of expertise behind, to some degree, in order to participate. Unfortunately this takes away our specialized skills safety net and increases our personal risk. I think for innovation, the question for IT types becomes, do we want to deal with that kind of risk?\u201d\nWe're onto something here. I disagree that \u201cmost IT people\u201d have been trained to be, or define themselves to be, idea killers (My loved ones include \u201cIT people\u201d who are among the most creative, innovative people I know). But the comment is right on about the need to accept more risk into your life. For example, I think there\u2019s frequently tension between the development organization that is building something new and cool and the operations group that has to integrate said new and cool thing into the application stack. The guys who have to keep the lights on need to be pretty risk averse. I'm not saying that\u2019s a bad thing because that SAP system has to keep humming. But the conflict is there, and it's not your classic alignment problem. It doesn't matter how much the CEO supports innovation, the core infrastructre and apps have to run reliably. So senior management has to resolve the tension and define when and how risk \u2013and failure-- is acceptable. \n\nJeneanne Rae recently had a good post that relates to this subject on businessweek.com. The post is about the conflict between Six Sigma and innovation. Rae observes that for an organization to support continuous improvement and innovation at the same time, it needs \u201cambidextrous\u201d managers who understand what\u2019s needed to support both activities. In other words, be able to manage processes that promote consistency and predictable outcomes, as well as those that encourage variation and risk-taking. The concept is from Charles O'Reilly and Michael Tushman, who wrote about \u201cThe Ambidextrous Organization\u201d a few years ago in the Harvard Business Review. (Subscription required)How do you develop managers within your IT organization who are comfortable with risk? Well, how do you get anyone to do anything? Set goals. Measure progress. Follow through. Unfortunately, a lot of companies don't do any of these things. A Boston Consulting Group study\u00a0published last summer found executives are increasingly dissatisfied with the return they get on their innovation investments. But the study also found that most companies measure very little when it comes to innovation and\u2013 even more important to my mind\u2014don\u2019t tie performance incentives to innovation. Among 269 executives interviewed, 47 percent said rewards of any kind are applied only inconsistently across projects, and 31 percent said they don\u2019t tie their innovation metrics to incentives at all. I\u2019ll bet that if the CIO\u2019s performance on innovation isn\u2019t being measured and rewarded, he isn\u2019t measuring the performance of his IT organization on that score, either.So now I\u2019m wondering: \u00a0What metrics do you think are effective for measuring the contribution IT makes to innovation? And how do those metrics roll up into your performance goals, and those you set for your managers?