By John Parkinson

Back in my college days, one of my professors gave me some advice that I have always remembered, but not always been able to act on. He told me, "Every experiment is a success if you learn something new from the results--whatever they are."

I've run a lot of experiments since then--and mostly I have remembered to learn something from the results. These days, however, it's not so easy to label each new learning event a success.

Corporate business culture all too often sees success as getting the outcome you expected (or promised, or hoped for...) rather than learning something new. This requirement for a known outcome to be validated flies in the face of experimentation, where you are trying to find out if your expectations will be met, not just validating them. (As Einstein famously remarked, "If we knew what we were doing, it wouldn't be called research.")

So last week I shut down a "successful" experiment. It taught us that the technology we were experimenting with wasn't quite ready for us yet--and that we weren't quite ready to take on the capabilities and potential of something new.

It hadn't cost much to find this out--and we learned enough through the experience to know that when the technology is ready, we will be ready too. A lot of people got to participate (which was good--they learned something) but some felt bent out of shape because we decided to pull the plug on something they liked, even though we weren't ready for a broader deployment. And quite a lot of people wanted to label the project a failure--simply because we decided not to continue the effort.

This is worrying. It's hard to innovate under any circumstances, but much harder if you have to be right the first time, every time. We need to recognize the value in an occasional "failure" like this and celebrate what we learned and the experience we gained.

Otherwise we won't be making much progress.

John Parkinson is CIO of TransUnion. Click here to read his columns for CIO Insight.