The Time of Webdesign, alistapart.com, published an article "In Defense of Eye Candy". Yes, I agree: people like nice looking things. I'd even go as far as to say that surveys, and studies don't necessarily do justice to how biased all of us are to visually pleasing things. I also think many people aren't aware of the real reasons why they might choose design A, over B. I have no doubt, that when it comes to design, the reality is as dubious as the results of this study.

The reason why I disagree with the article are:

The vast majority of projects [e.g. under 50k] put too much time into eye candy not too little. This time is rarely well spent.

With smaller projects, the questions that tend to need to be solved first are as follows:

1. Does it work?
2. Can I explain the purpose and value of the website to my 80 year old grandmother.
3. Is the messaging coherent, and worth reading?
4. Does the design succeed in not getting in the way of the main objectives of the website?
5. Can we actually built/adapt with the navigation/IA rules that we decided were important in our meetings?

Too often however, I've found that "what color should links be" stonewall questions like: "can people figure out what a link is?".

I've lost weeks, if not months of my life sitting in meetings debating the pro's of design A (with fake content), vs design b (also with fake content) only to launch a mediocre website with mediocre results. Turns out, gradients don't convert customers -- no loss though, good designers tend to be too expensive to actually give most companies a return on their investment. This is not to say "know your place." Rather, I'm saying, spend money on what makes you money.

In my humble opinion, the difference between the top 1.0% and the top 0.1 percent of websites is eyecandy. However, the difference between the top 50% and top 1.0% is all funciton, and communications, baby. So the cliche goes, no need to put lipstick on a pig.

One Rant:

As far as the importance of the look and feel of a submit button, I call 100% bullshit. There's no harm in showing users a form that looks like a form, in whatever archaic browser they happen to be using. Oh, and BTW -- good luck making that submit button work as well as a native one in all browsers you need to support. Really, I should bill my clients $400.00 for a submit button that is yellow with shadows? Get real, or offer me a job.