There are some good arguments here, namely: the quality of a site’s design has at least as much impact as usability on the site’s overall performance (that is, the site’s ability to support a business’s objectives by helping customers accomplish what they want to accomplish.)
Like any first impression, a bad visual design can undermine even the best usability, just as poor usability can undermine even the most brilliant visual design.
But there are some problematic statements in this article. First, the comment that “In such studies, participants have a set of specific tasks to accomplish, and thus their gaze tends to focus on navigation links, titles, labels, and interface controls such as buttons and form fields,” is misleading. Many studies like this are in fact either open-ended, or loosely directed; participants are asked what kinds of tasks they normally try to accomplish online, and then asked to do those things in the lab environment. In an eCommerce evaluation, for example, users are asked to purchase something they would normally purchase online, but to just do so in the lab. Setting up a counter-argument to Jakob Nielsen, though, is just too easy and for the author of this article, intellectually lazy; Neilsen intentionally writes with a bombastic style, mostly because the first 30 years of his publishing life were spent trying to get anyone to listen. (And besides, he never actually asserts that graphics have no effect on his participants’ experiences, he just points out that there is no evidence suggesting that the graphic elements make any meaningful contribution to one user’s outcomes or preferences.) Finally, it’s unfair to attempt to ferret out focus-group-like conclusions from a usability evaluation, even an eyetracking one. Eyetracking evaluations require only a dozen or so users in order to produce valid, reliable data; a study in which you wanted to evaluate attitudes and preferences would require hundreds of users in multiple locations in order to produce statistically-reliable conclusions. One reason to do this is to reduce the effects of groupthink and self-censorship that often rear their heads in focus groups.
Lips can deceive, but the eyes don’t lie. All of that is beside the point, though, because it plays into the well-worn meme of design vs. usability. I can understand the roots of this dichotomy (it’s another variation of the liberal arts vs. “hard” sciences debate), but I just don’t understand why it still persists. There are two sides to the same coin, and neither is sufficient to make or break the overall experience.