Thursday, March 31, 2005
the "beyond usability" debate
"Beyond usability" is a phrase one is starting to hear often. Google lists nearly 10,000 pages containing the phrase. What's going on? Is usability somehow inadequate?
In days past, people debated labels, such has how usability differed from user experience, or information architecture differed from interaction design. Other people drew elaborate diagrams trying to show how everything was different but still related. Two things have since happened: people got bored debating labels, and the definition of usability in particular has been definitively articulated through ISO standards. Now that we agree what usability is, can we live with it?
Usability has survived, and triumphed, over earlier criticism that it was barrier to "good" design, which usually meant the creative impulse of the designer. Now it is being challenged by more thoughtful questions about how to address innovation and the aspirational needs of users, and not just be about fixing things that are wrong. And even some usability professionals are wondering if being the critic is getting tiresome. An article last year in Interactions asked the profession: "Are you positive?", citing the need for human factors professionals to curb their critical dispositions. I've talked with others who have expressed feelings of existential tedium over the treadmill of just fixing stuff.
Caroline Jarrett, a seasoned usability professional and coauthor of a new HCI textbook, has explored the richness and limitations of usability in her article, Not beyond Usability - just nearby. Caroline notes that usability is supposed to cover satisfaction, but it is "a concept that's both static and slippery. 'Static' in the sense that it's a one-shot type of concept: you're either satisfied with whether the product allowed you to achieve your goals, or you're not. 'Slippery' in the sense that it can have so many meanings pushed into it: delight, enjoyment, a mild lack of discomfort, a major thrill." She questions if satisfaction (typically the stuff of likert scale questionnaires) really captures how engaging an interactive product is. She asks: "What emotions or other aspects of usability do you think we should add into the definition? And will satisfaction' encompass them? Or how do they fit in?"
The issue of usability's treatment of user engagement has created several opinion clusters. On one side are people who feel usability has never been in better shape, and there is nothing to fix. They are "usability engineers" and proud of it. In their view the profession is moving toward scientific respectability, with standards, databases of user responses for cross application comparisons, and common reporting formats. To measure emotion is fine, in theory, but you'd better make it scientific, not some namby-pamby nonsense.
On another side are a small group of researchers who think emotion can be incorporated into design using evidence-based recommendations. Examples are found in the edited volume: Pleasure with Products: Beyond Usability. Personally I find this work, which centers around conferences sponsored by the Design and Emotion Society, very interesting. Some fantastic insights have been developed so far, but we are a long way from truly understanding users' emotional needs, and even further from taking those insights and translating them into design recommendations. It is one thing to get inspiration from some lose or narrow research around emotions, but quite another to figure out if something needs fixing. The "design for emotion" field is burdened by the lack of a solid common understanding of human emotional needs in general, which needs to come from psychology and neuroscience. Emotions are very complex. (I personally despair of the cognitive science and AI types who venture into emotional research, muddying understanding with factless theory and computer simulations.)
In the absence of a methodical way to assess user emotional needs and develop solid recommendations to address them, still others are taking a "just do it" approach to designing for emotional needs. In the absence of a conceptual framework, the results are predictably random. On the positive side, participatory design can capture some emotional needs that suggest promising design solutions. A separate trend is the increasing co-mingling of branding with user research, which can appeal to companies interested in getting users more engaged with their product. As usability has gone mainstream, design agencies (often owned by ad agencies) have set up usability teams. And some usability firms have taken branding on board as part of the "user experience" to differentiate themselves from the increasingly crowded usability market. The extent to which brands are intrinsic concerns of users, emotionally or otherwise, is a separate debate that dwarfs "beyond usability" considerably.
For my part, I do think that usability needs to address emotional needs more systematically, and as far as possible, robustly. I am not sure if that will be possible within the measurement-focused orientation of the discipline. Even if one is not a quantitative tester (and most are not), most usability questions are asked in terms of binaries: does something work or not? Emotions are analog, and resist such treatment. I do not believe usability, as it exists in practice, is nearly as scientific as some would wish. What usability does offer when evaluating effectiveness, and can bring to user research on emotion, is a (intellectually) critical attitude. By definition emotions are subjective, but one can still probe them deeply and question what they mean. The usability field has a tradition of thoughtfulness that doesn't exist in fields like advertising, where is not uncommon for observers of focus groups to spend most of their attention munching on food and chatting while beyond the mirror. This thoughtfulness is especially true for those trained in broader, observational research techniques, such as grounded theory. On this basis, user research on emotional needs can be something grounded in evidence, and rather than based on whims and hunches, as so much market research can be.
In days past, people debated labels, such has how usability differed from user experience, or information architecture differed from interaction design. Other people drew elaborate diagrams trying to show how everything was different but still related. Two things have since happened: people got bored debating labels, and the definition of usability in particular has been definitively articulated through ISO standards. Now that we agree what usability is, can we live with it?
Usability has survived, and triumphed, over earlier criticism that it was barrier to "good" design, which usually meant the creative impulse of the designer. Now it is being challenged by more thoughtful questions about how to address innovation and the aspirational needs of users, and not just be about fixing things that are wrong. And even some usability professionals are wondering if being the critic is getting tiresome. An article last year in Interactions asked the profession: "Are you positive?", citing the need for human factors professionals to curb their critical dispositions. I've talked with others who have expressed feelings of existential tedium over the treadmill of just fixing stuff.
Caroline Jarrett, a seasoned usability professional and coauthor of a new HCI textbook, has explored the richness and limitations of usability in her article, Not beyond Usability - just nearby. Caroline notes that usability is supposed to cover satisfaction, but it is "a concept that's both static and slippery. 'Static' in the sense that it's a one-shot type of concept: you're either satisfied with whether the product allowed you to achieve your goals, or you're not. 'Slippery' in the sense that it can have so many meanings pushed into it: delight, enjoyment, a mild lack of discomfort, a major thrill." She questions if satisfaction (typically the stuff of likert scale questionnaires) really captures how engaging an interactive product is. She asks: "What emotions or other aspects of usability do you think we should add into the definition? And will satisfaction' encompass them? Or how do they fit in?"
The issue of usability's treatment of user engagement has created several opinion clusters. On one side are people who feel usability has never been in better shape, and there is nothing to fix. They are "usability engineers" and proud of it. In their view the profession is moving toward scientific respectability, with standards, databases of user responses for cross application comparisons, and common reporting formats. To measure emotion is fine, in theory, but you'd better make it scientific, not some namby-pamby nonsense.
On another side are a small group of researchers who think emotion can be incorporated into design using evidence-based recommendations. Examples are found in the edited volume: Pleasure with Products: Beyond Usability. Personally I find this work, which centers around conferences sponsored by the Design and Emotion Society, very interesting. Some fantastic insights have been developed so far, but we are a long way from truly understanding users' emotional needs, and even further from taking those insights and translating them into design recommendations. It is one thing to get inspiration from some lose or narrow research around emotions, but quite another to figure out if something needs fixing. The "design for emotion" field is burdened by the lack of a solid common understanding of human emotional needs in general, which needs to come from psychology and neuroscience. Emotions are very complex. (I personally despair of the cognitive science and AI types who venture into emotional research, muddying understanding with factless theory and computer simulations.)
In the absence of a methodical way to assess user emotional needs and develop solid recommendations to address them, still others are taking a "just do it" approach to designing for emotional needs. In the absence of a conceptual framework, the results are predictably random. On the positive side, participatory design can capture some emotional needs that suggest promising design solutions. A separate trend is the increasing co-mingling of branding with user research, which can appeal to companies interested in getting users more engaged with their product. As usability has gone mainstream, design agencies (often owned by ad agencies) have set up usability teams. And some usability firms have taken branding on board as part of the "user experience" to differentiate themselves from the increasingly crowded usability market. The extent to which brands are intrinsic concerns of users, emotionally or otherwise, is a separate debate that dwarfs "beyond usability" considerably.
For my part, I do think that usability needs to address emotional needs more systematically, and as far as possible, robustly. I am not sure if that will be possible within the measurement-focused orientation of the discipline. Even if one is not a quantitative tester (and most are not), most usability questions are asked in terms of binaries: does something work or not? Emotions are analog, and resist such treatment. I do not believe usability, as it exists in practice, is nearly as scientific as some would wish. What usability does offer when evaluating effectiveness, and can bring to user research on emotion, is a (intellectually) critical attitude. By definition emotions are subjective, but one can still probe them deeply and question what they mean. The usability field has a tradition of thoughtfulness that doesn't exist in fields like advertising, where is not uncommon for observers of focus groups to spend most of their attention munching on food and chatting while beyond the mirror. This thoughtfulness is especially true for those trained in broader, observational research techniques, such as grounded theory. On this basis, user research on emotional needs can be something grounded in evidence, and rather than based on whims and hunches, as so much market research can be.
Link to this article. posted by Michael Andrews @ 10:04 AM 
