I listened to a talk by the head of a prominent design school recently. He explained that the school's philosophy is to train students to analyze the needs of their users in order to design reasonable products. Ethnographies, focus groups, cognitive modeling: Interaction design, at least, has become a process dedicated to catering to the user. It seems as though most designers and educators agree that user-centered design is simply the right way to do things, and a recent study confirms its pervasiveness . It's almost a religion in our design schools, but it is in fact an incomplete philosophy that lacks a sense of responsibility for concerns other than those of the immediate end user.
Although we might wish for word processors and faucets that are more intuitive, as we look around in the world, it's easy to see daunting, more serious problems: polluted water, war, excessive atmospheric carbon dioxide, a pressing economic crisis. There's no question that these issues are, in general, the most serious challenges facing us today. They threaten our very existence. But designers have made scant headway in intervening and creating solutions to these sorts of problems, and I believe it's because of a misguided focus on the user. An example may help to illustrate.
Imagine a designer tasked to improve a store's shopping-bag system. Years ago, a big company might have focused on reducing costs as the No. 1 criterion for a design task like this, but imagine that for our task, the company has chosen a modern, user-centric approach in the redesign of its bag system. Realizing, perhaps, that a bag that meets the user's needs is more likely to encourage repeat business, the designer immerses herself in the experience of the store's customers. She conducts interviews, assembles cultural probes, and does targeted user observation. It wouldn't be at all surprising if the final design were a thick, disposable plastic bag. It would meet the user's needs of being inconspicuous, robust, lightweight, and even comfortable to hold when loaded.
The problem here, of course, is that disposable plastic bags are clogging landfills and creating air pollution through their manufacture. These two (and there are many more) adverse side effects of the design solutions are far bigger problems than those experienced by a typical retail customer at checkout. Our user-centered design process has answered questions of convenience, comfort, and cost, but created problems that are far more serious. It could be argued that the designer's analysis of the user's needs was incomplete. After all, no retail consumer really wants poisonous air or plastic bags littering mountain streams. We might say user needs must be more explicitly and thoughtfully derived, and that we can attempt to satisfy these bigger, more critical needs if we can get deeper into the minds of users. But is that true? There's a simple solution users can effect on their own: bringing their own reusable bag. People aren't stupid; they know this is a great way to make a little dent in a large problems. But nobody brings their own bag to Wal-Mart. Smart design decisions can have tremendous influence and can effect change far faster than the gradual behavior change of individuals.
A popular example of good user-centered design is the OXO Good Grips line of cooking utensils. It is successful on many levels: The utensils are easier for children and people with arthritis to manipulate, and they have succeeded wildly in the marketplace. This is, without a doubt, an example of great design success. OXO wanted to make the most money possible, of course, and the design firm (Smart Design) provided a solution that is appealing to many. But this universal design process results in products for the lowest common denominator of user, and now everybody, not only the elderly, ends up with a can opener encased in a big plastic shell. Santoprene, the soft material used in the product, is easily recyclable, but OXO (according to customer service) does not use recycled material in their products. When the number of items sold runs into the millions, this begins to seem like a significant waste.
Environmental concerns are clearly a dominant theme here, and designers have recently started thoughtful and promising initiatives like the Designers' Accord  to address problems of sustainability in design, while theorists and educators have raised these issues in the literature with greater frequency [3, 4]. But my criticism is more than a call for greater sensitivity to the environment; it is also an acknowledgement that reliance on our understanding of our users' needs has gotten us into this mess. In her call for papers for a special journal issue on user-centered design, Anne-Marie Willis explicitly raises the unanswered question of whether user-centeredness and conservation are at odds .
Another obvious example is the SUV. People wanted more interior room and a more commanding position on the road, so the cars became larger and larger. Not only do SUVs pollute the atmosphere at an alarming rate, but they also make it more dangerous for the rest of the people on the road. Pedestrians and small-car passengers are no match for a 7,000-pound truck. Wanting to protect your passengers in the event of an accident is understandable, but a valuation of the worth of human lives is not something that should take place on the floor of a car dealership. It's 2009; we understand this now. But what about Detroit? It seems they built their businesses around users' immediate desires with no planning for the future. And now we are all paying the price.
SUVs and Wal-Mart are obvious examples of off-center design. But to bring the discussion closer to home, consider the desktop printer software I installed yesterday. Its interface is probably the result of thoughtful user testing. I imagine the interaction designers compiled data reflecting the most often used settings and set the defaults accordingly, which seems reasonable. But the defaults are for single-sided printing at high quality, using more paper and ink than is probably necessary for most tasks. On the operating system that I am using, those defaults cannot be changed.
Centering design decisions around what a user wants, or even what a user needs, is misguided. Giving precedence to a single person, or group of people, instead of taking everything else into account as well, is the root of our major problems. Side effects from poor design decisions are killing us. This might be obvious on a political, social, or even economic level, but it is every bit as germane in design.
As interaction designers, we often intend to support the behavior of our users as well as possible. We might feel that it's not our place to suggest duplex printing, that decisions like that will eventually be motivated by cost or by law or by, really, someone else. And at that point, we can reflect the change and support the changed view. But this just marginalizes us. There is nothing wrong with trying to change behavior. That's what designers are supposed to do.
How can we do it? I believe that it's up to us, the design community, to articulate this change. As a group, we can reduce the emphasis on the user and broaden the scope of our process. We need a new direction to guide our work and to educate our students in design schools. Don Norman has proposed "activity-centered design" , but this is similar to the current philosophy, with more context taken into account. Inspiring work by Cameron Sinclair and Kate Stohr might suggest "humanity-centered design" , but that sounds like we're ignoring the welfare of sea turtles. We could try "decentered design" or, "centered design," but instead of more buzzwords, I vote simply for a concerted effort to take a broader view in our design process. To take into account the effects of our decisions on non-users, secondhand users, animals, and the Earth. To take a wider-eyed, equitable look at design problems and to take efficiency, sustainability, and public safety into account at every design decision.
It's great to make things usable. I'm frustrated trying to open clamshell packaging or redeeming frequent flier miles on a poorly designed Web page. But many decisions are not win-win; a gain for a single user is often a loss for others. Design is not all about ease of use and convenience. We need to determine when to make things difficult or unpleasant for users. We must question the assignments we work on to see whether better problem formulations exist. This is difficult. But it is the type of work that designers, more than anyone else, are capable of doing. Designers are skilled at working on multiple, concurrent solutions. We are skilled at taking the views of multiple stakeholders into account. We are often good at thinking about tangential effects of our work, of unintended uses and circumstances.
This call to expand our focus just might not resonate for some interaction designers. The layout of a GUI might have tremendous influence on an end user's productivity, and the placement of buttons might not affect anything else. But our current focus on the user didn't originate with these stereotypical interaction design tasksthey were introduced in the 1950s by Henry Dreyfuss  and popularized by Don Norman . An oft-cited example in Norman's book is of his lab's experiment to improve the usability of their lighting system by rewiring so that the lights are controlled by a bank of centralized switches, laid out on a diagram of the floor plan. This makes sense, but is getting around the inconvenience of having to flip a switch on and off (or of learning which switch goes with which light) worth several hundred feet of copper and PVC wiring? With all things equal, I'm in favor of making things easier to use, but most of the time, all things are not equal. One might respond to this criticism that the book was written before concerns about waste became so urgent. But more likely, the response would be the research was focused only on usability issues, ignoring other factors like cost and material waste. That is precisely the problem.
User-centered design is wrong. But the current myopic view of a designer's responsibilities is not anyone's fault. I don't believe people make shortsighted decisions out of laziness, but because they lack the appropriate tools and information to make better ones. It's not trivial, for example, for designers to accurately predict the side effects of using plastic or paper wrapping on a product because the network of impacts and stakeholders involved in these decisions can be mind-numbingly dense. Not only do we suffer from a lack of design theory that takes emergent, complex systems into account, but we also lack solid analytical theories of these systems. Life cycle assessments are valuable tools for understanding long-term effects, but they are complicated processes and not accessible to designers making everyday decisions. We will certainly see new tools that can help us make more informed decisions, but it would be a long stretch to imagine that we will be able to use them as a basis for a prescriptive theory of design.
Engineering offers "Design for X" as a way of managing the complexity of design problems, where X represents one of the "-ilities." These are items in a long list of quality attributes (accessibility, reliability, usability) that engineers must take into account in their designs. Due to the size and scope of some projects, engineers need to switch focus between different "-ilities," improving certain aspects of their design while balancing each particular goal. Keeping multiple objectives in mind without centering on a single one leads us to a promising model for interaction design. The distinction is important: "Design for usability" must be in our set of "-ilities," but centering design problems on the user leads to off-center solutions.
A call to bring balance into design is nothing new. Buckminster Fuller is widely credited as a genius, and his designs were notable for effecting broad change, even if they were ahead of their time. He called himself a "deliberate comprehensivist" and advocated a broad view, questioning the formulation of design problems. Even Don Norman writes, "If everyday design were ruled by aesthetics, life might be more pleasing to the eye but less comfortable; if ruled by usability it might be more comfortable but uglier. If cost or ease of manufacture dominated, products might not be attractive, functional, or durable. Clearly, each consideration has its place. Trouble occurs when one dominates all the others" .
These shifts in the focus of design may seem cyclic. Many of the poor designs that Norman identifies in The Design of Everyday Things exist because of a need to save money. Now, after a period of intense focus on the user, it may appear that I am advocating a return to the old days of cutting costs and materials. But this is not the whole story. I am advocating a balanced design process that considers as many factors as possible instead of focusing foremost on the end user's needs. Design is by nature a series of trade-offs, and while every situation is unique, always trading in favor of the user is rarely a smart idea.
Warm thanks to Eli Blevis, Dan Boyarski, Mark D. Gross, and Gabe Johnson for their comments and conversations on this controversial subject.
Eric Schweikardt is the designer of roBlocks, a modular robotic construction kit for education. He is also a visiting scientist at the Computational Synthesis Lab in Mechanical and Aerospace Engineering at Cornell University. Schweikardt recently completed a Ph.D. in computational design from the School of Architecture at Carnegie Mellon University. His current research looks at complex, concurrent systems and how our notions of design change when working with thousands and thousands of tiny robots.
©2009 ACM 1072-5220/09/0500 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.