The usability movement is sometimes criticized for being dull and for promoting boringly invariable designs. The chief reason for this is that some people equate design conventions with creative restrictions. However, this equation doesn't add up for two reasons.
Standards vs. Design Dictates
First, while it is true that usability is typically enhanced by consistency and adherence to design guidelines, this does not necessitate identical design. Rather, such conventions aim to create a vocabulary of building blocks that designers can combine in many vastly different, and often enjoyable, ways.
Consider natural language. Each word has an established meaning, and we typically combine words using a defined grammar. Literature that follows these conventions is easier to read and has a bigger audience than avant-garde, experimental literature. Still, such "conventional" novels are definitely not the same: Although they use fully standardized language, they can reach any desired extreme on a variety of emotional scales.
Usability = Engagement
The second reason usability is not opposed to fun is that the greatest joy of using computers comes through user empowerment and engagement. It's very enjoyable to visit a website that works, where everything just clicks for you. In contrast, a user interface that doesn't do things the way you want feels sluggish, unpleasant, and possibly even hostile, despite the designer's no doubt sincere attempt to invoke positive emotions. A user's personal experience trumps anything the designer is trying to communicate. In talking about a design's "look and feel," feel wins every time.
As an example, Amazon.com uses associative links to create a fun and rewarding experience for users. Each book page offers associative links to five books frequently bought by other people who purchased the book youre interested in. Following these links can lead to a powerful feeling of discovery. As a result, you can easily spend much more time shopping on Amazon than is dictated by the simple efficiency metric of buying the book you came for as quickly as possible.
Such engagement requires usability. If users can't master the interface, they'll feel oppressed rather than empowered, and are unlikely to explore or use anything beyond the absolute minimum. On the Web, this "minimum" often turns out to be one or two page views, and then users are gone -- never to return.
There is certainly more to an enjoyable activity than the mere ability to complete it. At the same time, computers are currently difficult to use and much of the Web feels like a vast wasteland. Given this, people can and do derive considerable pleasure in finding a well-crafted user experience that empowers and engages them.
Methods for Testing Satisfaction
Traditional user testing is great at debugging user interface designs to find the elements that make the system difficult to use. Test methods are less evolved, however, when it comes to determining the enjoyable aspects of a design. In the past, this was not much of a problem because user interfaces were so difficult to use that all we could hope for was that they'd improve to the level where using them was not actively unpleasant . Websites in particular were designed in such great contrast to users' needs that simply exterminating bloated and useless designs has been the usability movement's great achievement over the last ten years.
Now, as we change from the negative endeavor of removing bad design to the positive pursuit of good design, we must modify the methodology to encompass more awareness of fulfilling, engaging, and fun design elements.
Most studies currently rely on classic and not completely satisfactory ways of assessing user enjoyment:
A subjective satisfaction questionnaire administered at the end of a study that provides a simple, overall system assessment.
Observations of the user's body language for indications of satisfaction or displeasure (smiles or frowns), as well as for laughs, grunts, or explicit statements such as "cool" or "boring."
A skilled observer can gain much insight from the second approach, but it is a weak and possibly misleading source of data for less-skilled usability professionals, who constitute the vast majority of the world's test facilitators.
As for the first approach, subjective satisfaction questionnaires suffer the standard problem of being administered out of context: They typically rely on users' recollection of enjoyment, rather than the actual experience of use in the moment. You can alleviate (though not eliminate) this by administering several small questionnaires throughout the test session rather than saving all the questions for one larger questionnaire at the end.
Beyond Ease of Use
As always, you cannot rely on simple, literal interpretations of users' statements. For example, in testing company websites, users almost always say that they don't want fun or entertaining content: Just give me the answers as straight and as fast as possible. And, in observing actual user behavior, we certainly do see negative reactions to frivolous content -- such as big photos of glamorous models or meaningless animations that bounce around the screen. But, at the same time, we also see users smile or exhibit other positive body language when they come across cleverly written content or moderately funny descriptions -- assuming they fall within the scope of users' expectations of professional writing in the website's genre. Thus, users seem to appreciate and enjoy a somewhat higher style to their content than they claim to prefer.
We need much better methods for testing enjoyable aspects of user interfaces. Such methods should be both robust and easy to apply, since people with relatively little expertise do the vast majority of user testing in the world.
That said, ease of use must remain our first priority. Technology is just too difficult for us to abandon this goal. But hopefully it will soon be time to emphasize joy of use as well.