A new Jacob Nielson article reveals eye-tracking results for recent banner ad studies. As should be expected, there is some really valuable stuff in there.

Obviously banner ads do work. People actually do click them. MySpace wouldn't have sold for $580 million if they didn't work. The real issue is that banner ads aren't for simply looking at. The are for taking you somewhere else. You don't spend as much time looking at something you are clicking as you do looking at something you are reading, just like you tend to spend more time sitting on the couch than you do walking through a doorway.