Only a little while ago, every webdesigners go-to knowledgebase — A List Apart — published: In Defense of Eye Candy. The article, on the surface, challenges the conception that eye candy for the sake of eye candy is not only okay, but it is more successful than the eye-candy-free interface. I was asked my opinion of this and I disagree (as avid readers of this yarnball of rants would expect). If you wish to skip the eye candy in this article, the conclusion is as it usually is these days: Google’s website is eye-candy free, and they do pretty well.
Going indepth, things are far more complex. More complex than the ALA article indicates even. It is a good article, it’s got compelling arguments and good general design advice. But it is also simplistic. The crux of the article is a quote from a 12 year old study:
Researchers in Japan setup two ATMs, “identical in function, the number of buttons, and how they worked.” The only difference was that one machine’s buttons and screens were arranged more attractively than the other. In both Japan and Israel (where this study was repeated) researchers observed that subjects encountered fewer difficulties with the more attractive machine. The attractive machine actually worked better.
These are interesting results. Unfortunately, they do not directly translate to screen-design, which is what ALA is mostly about. I suspect the success of this particular ALA article is due to its pseudo-sensationalist headline. Let’s dig deeper.
Eye candy refers to visual elements that taste sweet, when eaten by an eye. As with all other terms, eye candy has been typecast to represent only that; meaning it’s never about usability or function at all. The ALA article states the obvious: eye candy doesn’t have to mean at the cost of usability. Somehow, however, I get the generally opposite vibe from the article; that usability is usually at the cost of eye candy. Which is not true.
Good designers create usable interfaces that look as good as they function. And they do more. A good designer considers audience and usage situations. A good designer considers consistency and conventions. A good designer knows that eye candy is to userfriendly interfaces, what the trailer is to the movie; certainly part of it, but not the main course either. I think the author of the ALA article, Stephen, knows this, but I also think he knew they eye-candy title would get more attention.
The example with the ATM machine is delicious to bring up, but also near useless. Working an ATM machine, your goal is to get your money out of the bank quickly and securely (which — for a designer — is certainly a usability challenge in its own right). But because an ATM machine works better when glossed over, it doesn’t also mean that you can now go ahead and skin the form-buttons on your website. Because the ATM is not your day to day computer, it’s a totally different situation, mindset and metaphor. Your day to day computer is the one your mom is scared of breaking just by turning on.
You are not the designer of the operating system that runs on your computer (if you were, skinning push buttons for cross-OS consistency would be an altogether different matter). As a screen designer, your job is to design an interface that is user friendly and attractive. So does that mean you get to design a glossy, bevelled drop shadowed button just because Stephen says it matches the physical metaphor? No, and it doesn’t mean you should make it flat and bordered either. It means that for each unique situation, you have to decide what’s best for the user, taking numerous things into account. Sometimes, Stephens yellow button is okay. Other times, especially in intranets and backends, the best thing to do (unless you have a really good excuse) is to not design the button and have the users operating system decide the look. As my argument for this has always been: it’s about consistency — when a button looks like every other button on the OS, chances are your visitors knows what it does. A yellow, bevelled button with a drop shadow is more likely to look like a graphic than a button because the user is unlikely to have seen it anywhere else (unless they just came from ALA where it is, ironically, also a graphic).
The bottom line is: the physical metaphor is not some goddess of usability which always translates to your computer interface. As Stephen says himself, aesthetics are about all the senses; sight, tactile feedback, smells, the lot. But unless your computer is on fire or from the future, you’re limited by the same flat screen as I am. That means, even if it looks like a button, it isn’t actually bevelled. Even if it does have a drop shadow, it doesn’t actually have substance. Even though it has a glossy gradient, it doesn’t mean the sun is at zenith. Which essentially rips the proverbial carpet away from the bulk of Stephens otherwise excellent article.
A while ago, I wrote:
In the end, what we think is pretty will fade or change given time. What was pretty twenty years ago might not be pretty today. Mullets come and go like the tides, so learn to spot the mullet and try to avoid it.
Pinstripes were the mullets that MacOSX wore in 2001.
Pretty really is relative. Back in the Bauhaus days or even the early nineties, people would probably have preferred the flat and bordered button to the glossy, rounded and graded button. In the future, they might again. If you’re blind, you’ll probably prefer the button that’s not a graphic because your screen reader will be able to read what it says. Oh, if only the physical metaphor always worked.
I predict I will be writing about this again next year.