Why Usability Folks Don’t Focus on Visuals in Design

In which I give several reasons why the usability vs. design debate is the wrong debate, and why we should be looking at why usability folks don’t focus on visuals when designing, like many designers do.

Sometimes, in an effort to discuss issues easier, we make them either/or. Either you’re for it or against it. No in-between allowed.

We do this for all sorts of issues, from politics to religion to personality. We talk of being either “conservative” or “liberal”, as if they are mutually exclusive and everyone isn’t a little bit of both. We talk about Christianity and Islam as if there aren’t countless variations of practice within each. We talk about someone being either an “extrovert” or an “introvert”, as if introverts never talk and extroverts never stop.

When it doesn’t work out, though, when a trivial split doesn’t accurately reflect the actual situation, we call it a false dichotomy. The black and white approach just doesn’t work – we must investigate the gray areas to find out the truth.

Usability Vs. Design, Again

This is what is happening in the usability vs. design debate, recently re-ignited in two articles by Dirk Knemeyer on Digital-Web. (I wrote about the first of these articles here). Dirk claims that the “culture” of usability is dying, and the pendulum is swinging back toward design, as if it ever left it. This sets up a dichotomy that isn’t real. Usability has no more “stunted design creativity” than design has stunted ease-of-use. Design is what we have to do to create something and each thing we create has some level of inherent ease-of-use. The field of usability simply offers us techniques to evaluate our design and refine that ease-of-use, if we so choose.

As is often the case with such controversial articles, the comments were as enlightening as the article itself. Both sides of the argument railed against each other, with one saying that “usability” is more important and should be the focus of design and the other saying that “conceptual, creative, and visual aspects” of design are more important and should be the focus.

I don’t think one side being more important than the other is the real issue here. I don’t think we need to “return to design”. All the usability folks I know are working hard on designs, not against them. However, though I don’t agree with Dirk’s main argument, I think he (and those who wrote comments) helped shed light on what is really going on: that unlike designers, usability folks don’t trust visual design as the primary way to make something easy to use.

Why Usability Folks Don’t Trust Visuals

Why is this the case? Why don’t usability folks trust visuals as the primary way to make something easy to use, like many designers do? This is the real crux of the issue: it certainly upsets designers when someone suggests that web sites don’t have to look good to be good. And it certainly upsets usability folks when the opposite occurs.

Here are several reasons why usability folks don’t trust visuals:

    Visuals Are Rarely the Goal of Users

    There are several reasons why someone might come to your web site to look at visuals. One that is familiar to me is when a web designer goes out looking for inspiration by checking out what their peers are doing. Another is when your site exists primarily to promote its visuals, be it a portfolio site or something really unique like Dave Shea’s CSS Zen Garden. And there are others.

    For the most part, though, looking at visuals isn’t the main goal of users. Instead, they’re focused on achieving a goal that has everything to do with the written content on the site. They might be looking up medical information, purchasing a book, doing their banking, reading the news, or responding to a blog entry. Mostly, they’re reading.

    Visuals Are Rarely A Failure Point for Users

    Reading is an interesting case. Even though visuals often aren’t the goal of users, users interact with them at every moment. Reading words is interacting with the typography of a site: groups of visual shapes that users interpret as words and characters, giving them meaning. In this way visuals (in the form of typography) are definitely part of every success and failure of users.

    Visuals are very important: they can help or hinder a user’s comprehension of the information on the site. This would seem to be a prime area for usability folks to evaluate on sites. However, it turns out that users don’t often fail to do what they want because visuals have failed them, they usually fail for some other reason. Most of the reasons for their failure have to do with content issues that are specific to their context. Common examples include: search results for their query were poor, they didn’t understand something, the descriptions of products they were shopping for weren’t detailed enough, the content didn’t answer their question, or they were simply in a hurry. Most designers would like to think that their visuals are crucial to a user succeeding or failing, but for the most part users find something if it’s on a page despite what the visuals look like. (Whether what the user needs to know is on the page or not is the more serious issue.)

    There is No Good Way to Measure Visuals

    As I said, although visuals are rarely a failure point for users, they are still very important to their success. Unfortunately, visuals are open to a wide variety of interpretation. Even though we know they affect branding, perception, and experience (as Dirk pointed out), we’re still not in agreement about the extent or depth of such interactions. One person’s art is another person’s ugh. In order to understand visuals and to take advantage of them in design, we need a good way to measure them.

    Unfortunately, we don’t have that � yet. As with other exercises that are highly subjective (like advertising) it is very hard to know what works and what doesn’t work with visuals. Is there a way to tell if the visuals are what is leading to the success of a web site? Did a change in visuals improve the site, or make it worse? It’s often just too hard to tell. If you can tell, you’ll be a popular person in the usability industry!

    Usability folks feel safest with things they can measure more reliably than visuals. Things like revenue, success rate, or number of complaints. Based on use, these metrics are easier to relate to success than visuals are. It is unfortunate, though, because it is obvious that visuals are important, we just don’t have a good way to measure how important.

    Evidence Suggests that Experience Shapes Evaluation

    Including tests we’ve conducted at work and/or heard about elsewhere, a few usability studies have suggested that a user’s perception of a web site is highly influenced by their success. In other words, users who succeed on a web site are much more likely to rate the site higher on a plethora of subjective criteria, including visual design. If they don’t succeed, they are more likely to rate the site much lower, even on criteria that had very little influence on their problem. We’re only beginning to understand this phenomenon.

    At first glance, though, this isn’t surprising. Those things we know and love have often done something for us to earn that love. They have been easy to use, reliable, or surprisingly functional. My experiences with amazon.com have all been positive, and I would probably score them relatively high on their visuals, even though they aren’t that attractive. But, of course, that’s not what I’m there for.

Visuals Are Important!

I’ve tried to give a few reasons why usability folks (Ok, I’ll speak for myself) don’t trust visual design as the best way to ensure ease-of-use. There are simply more easily-measured variables on web sites than visual ones.

Notice, however, that I’m trying hard not to discredit visuals here. I’m simply suggesting that we don’t yet know how important their role is in design other than to say “pretty darn important”. For many usability-minded folks, that is enough to rail against articles such as Dirk’s when he says that “usability culture must end”.

Furthermore, research also suggests that pretty things may be more usable than ugly things. Noam Tractinsky, an Israeli scientist, published What is Beautiful is Usable in 2000, based on a study (popularized by Donald Norman in Emotional Design) of two designs that were identical except that one was pleasing to look at and one wasn’t pleasing to look at. The one that was pleasing to look at scored better in perceived ease-of-use, with actual ease-of-use having no effect. If the results are true, this study is another indication that perceived ease-of-use is distinct from actual ease-of-use (also suggested above), and users are swayed by visuals when rating ease-of-use. Currently, Tractinsky is working on ways to measure aesthetics in design, which would certainly serve to make this debate even more interesting than it already is.

So, usability folks (as a culture) aren’t against design or designers in any way: that’s a false dichotomy that might lead us down the wrong road. Instead, usability folks are unsure of how to assess visuals compared to their more concrete counterparts.

As for design, onward and upward. In no way am I trying to lessen the value of visuals in design: I’m simply offering an explanation for the usability side of things. There is no design in the world that shouldn’t be pleasing to look at, clear in its display of content, or inspirational to others.

As one commenter to Dirk’s article pointed out, we do have the same goals, after all. That goal is to make great things for others to use and enjoy. I think if we look at these issues and discuss them as challenges instead of taking sides and treading on the idea of who’s work is more important, we’ll end up with a clearer picture of where to go next.

Then, we’ll be working together instead of drifting apart.

Published: November 23rd, 2004