I have talked about usability testing for some time, but the talk got bogged down in looking for supporting technology, and thinking I wanted statistical validation.
Matthew shared a link with me that is interesting, and I think is right on: testing with one person is 100% better than no usability testing.