I have overall mixed feelings about this "user test." Maybe it's just a problem with what the industry calls "user tests." The demand characteristics of the tasks in this "test" make it difficult to make any real conclusions from the data. It's not measuring real user behaviour.<p>The tasks were great for identifying some confusing elements, but this was more or less a focus group and barely analyzes true usability issues. User preference or evaluation != usability problems. The only thing I found useful was the feedback regarding the menu structure. I did think the "click where you ____" was a good way to gather feedback, though.
Well at first glance it's interesting to see that the TC logo gets so much attention/clicks.<p>It's as if people are thinking "it should do something", or "wtf is that?".<p>All in all though, this redesign must be some kind of prank...
Their tasks:<p>1. Click on the things that draw your attention the most.<p>2. Click on the elements you like on this page.<p>3. Mark the things that you think we should improve.<p>These are the wrong tasks to be asking users to complete. A real user task on TechCrunch would be "Find and read an article that seems interesting to you" or "Find and read the 3 most recent articles about FourSquare" and so on.<p>This is basically a focus group for "do you like this redesign", not a task-based usability test.