So, why is that such an awful example for us?

In IT, we need to listen to users too. To be successful, we have to pay attention to what users actually do with the software and hardware we deploy. Those people are the ones who can tell us what's right and wrong with our systems, and show us problems and advantages we'd otherwise never know about.

But we don't have the luxury of years to build applications that bundle together a lot of good ideas. We can't afford an identity crisis. And we definitely can't risk waiting until after we've dropped a system in users' hands to find out whether we're on the right track.

Our job is to cut that risk -- with prototypes, user input and feedback, and plenty of hands-on testing as we develop a system. Sure, we know what works. That's not enough. We need to learn what our users will do with it. And we need to learn that early and often.

We have the opportunity to know what our users need before we finish our elegant, stylish work. And we can't afford not to.