About Face 2.0(c) The Essentials of Interaction Design

The hype that surrounded the Web and its concomitant New Economy led to many popular myths concerning the design of Web content and Web-based applications. We can forgive this kind of mythology, which is typical of most popular technical innovations. When commercial and industrial use of electricity became a reality at the end of the nineteenth century, it was commonly believed to have miraculous powers not only to transform business and society, but also to cure almost any conceivable ailment by "restoring the life force." Like electricity, the Web has transformed our society to some degree, even in the relatively short time span since its introduction. However, it took 50 years for our use of electricity to mature, and it will doubtless take as many years for us to realize the full benefits and socioeconomic ramifications of the Web.

Some of the most common myths about Web design follow. These myths have found their way into business and technical organizations, and are — to some degree or other — taken at face value by management, marketing, engineering, and sometimes even Web designers themselves. The sooner you can disabuse your organization of these myths, the better.

• Myth #1: The Web inherently makes things easier to use. In the early days of the Web, all that was possible in terms of interaction was single-clicking on text or images to move to a new page (or another part of the same page). This incredible dearth of interaction severely limited what was possible to implement on the Web, with the side effect of making behavior extremely simple. In this sense, the early Web was easy to use, although poor structure and layout could still easily ruin the user experience. However, with the subsequent addition of features like frames, forms, scripted actions, and embedded applets, these apparent gains in ease of use have vanished. What designers are left with now are systems that appear rich enough to be desktop applications. For lingering technical reasons, however, they are capable neither of supporting all the idioms we have come to expect of desktop applications nor of providing the smooth flow and rich feedback that desktop software allows. Dealing with these issues remains a substantial challenge for Web-design practitioners.

• Myth #2: Designing for the Web is new and different. For the many print-oriented graphic designers who entered the Web-design world in the mid-1990s, the Web was certainly new and different. It was more constrained, more fluid, and it introduced a new (to them) concept: hyperlinking of documents. However, little of this was news to software designers, GUI programmers, and usability professionals, who had been concerned with nearly identical problems in the design of software interfaces for many years. One area that was new was the emphasis on content over form and behavior, and thus the discipline of information architecture quickly grew to address that gap in knowledge. However, as the cutting edge of design for the Web has shifted from presentation to transaction, behavior is once again the dominant concern, although content must remain quite important as well. The process of design for the Web is not really different from the design of software applications, although the requirements of content and certain technical limitations affect the direction that the design solution ultimately takes.

• Myth #3: Web design is about HTML, layout, and typography. Since the beginning of the Web, there has been an unfortunate terminology problem in Web-design circles. Print and graphic designers, moving to the Web, brought with them a media-focused perspective on design. Traditional art and design schools focus considerable attention on mastery of media, and rightly so. However, this media-oriented way of thinking about design applies best to design problems that have little or no complex structure, no architecture to worry about. Jumping directly to the crafting of individual pages short-circuits the important planning phase that information architects soon recognized was required for Web sites more complex than a handful of pages. Again, this mirrors the experience of designers in the software-application world, where planning and prototyping is the primary design activity and where implementation is, subsequently, performed by programmers. Because Web implementation has historically been relatively easy at the front end, it has often been lumped together with the design process. For many graphic designers, who were recruited for Web design in the boom times, the design of a graphic image also means its implementation. However, to produce superior solutions, most commercial Web sites today require a greater division of labor between design/planning and implementation.

• Myth #4: Web design is just about the front end. As we have hoped to show in other sections of this book, design concerns the whole of the product or service — not just what it does, but what it is and what goals it serves. This is no different for the Web; the design of the user's experience, her interaction with and perception of the system, implies that the system must be considered as a whole: how form, content, and behavior are presented by the front end, and what those behaviors must imply at the back end. Web design is concerned with visual design, information design, and interaction design, in near equal parts. The latter two, in particular, can have significant ramifications for the design of the back end of a transactional system.

• Myth #5: Web design is about browsers. Although the majority of Web applications currently run inside of browsers, change is already in the air. Internet-enabled software applications have existed for a long time, and with the advent of technologies like .NET, the number of Internet-enabled programs existing outside of browsers will rapidly expand. Besides desktop applications, device-based applications that incorporate wireless connectivity are already beginning to gain popularity as products like the Handspring Treo become available. Web-enabled software of any variety promises to deliver on the same kind of rich information and media access the Web provides without the constraints of browser-based interfaces. It is up to interaction designers and the product team to decide what platform is the most appropriate for a new product or service: browser or standalone application. Browsers work reasonably well for browsing. Complex transactions are better handled by more sophisticated interaction models with richer idioms at their disposal.

• Myth #6: Web design is about Web pages. Before the popularity of e-commerce, Web design was very much about page design. Of course, even then, it was also about overall site structure, logical flow, and navigation. The view of Web sites as hierarchies (and other-directed graphs) of reasonably static pages persists today, and this notion is reinforced by a body of work published about the discipline of information architecture. Clearly, there is still a place today for thinking about informational Web sites from the perspective of the page. This paradigm is, at the same time, fundamentally at odds with the way most highly-transactional sites are generated — dynamically from information retrieved from databases in real time. As front-end Web technology gets more sophisticated interaction capability, it makes less and less sense to treat dynamically generated screens as pages. Highly transactional Web applications are best considered as what they really are: client-server applications. Breaking out of the constraints of page-oriented thinking allows designers to better address the real goals and needs of users of transactional Web applications.

• Myth #7: The Web and the Internet are synonymous. Early Web technologies (HTTP and HTML, in particular) were designed for transferring and formatting ASCII text and the occasional binary image in an asynchronous, request-driven manner. Web browsers and servers remain well suited to these kinds of operations, but not as well suited to transferring large amounts of heterogeneous data in a synchronous or streaming fashion. The Internet is a much more diverse network than the browser-accessible Web. Distributed applications that use Internet protocols (IP, TCP) can be much more flexible and powerful than browser-based Web applications.

• Myth #8: Web applications are easier and faster to build than native desktop applications. HTML allows for fast prototyping and some information-oriented Web sites have simple behavior that permits rapid construction. However, a transactional Web application with complex behavior entails the same kind of engineering challenge as the development of native code, and inevitably it needs to be approached using a similarly robust design and development methodology.