USC Annenberg Online Journalism ReviewUSC

Publishing by Design: Time to Make Human Factors a Concern


Digital communicators let fallacies about convergence and traditional media dominate their thinking. The relatively new field of human-computer interaction (HCI) can offer alternatives.

Two fallacies dominate the world of digital mass communication: 1. That media are converging into a single device. (Stated explicitly) 2. That pre-digital models of communication can be extensively used for specific digital purposes. (Implicitly tolerated)

The first fallacy, the concept of convergence, is largely a result of misinterpretation. Convergence has taken place at the back end of Internet technology, at the level of data formats and protocols, at the level that things are made and distributed. At the front end, people need divergence.

New media publishers have neglected problems and solutions at this critical first stage where people choose a device to access or deliver news and information. The field of human-computer interaction (HCI), which studies our cognitive and physical interaction with computing devices, addresses many of these issues.

Here is one axiom grounded in HCI: The personal computer/Web browser model does not represent the acme of mass communication. It is just one way in which people can find and review published material, and an inconvenient one at that. PCs are expensive and difficult to share; they are typically bulky, inelegant and immobile; they take time to start up and many broadband Internet connections are not even "always on"; Web browsers are flaky and only really capable of showing "pictures of data." Most of all, PCs are just plain hard to use. They are a universal tool, a jack of all functions and master of none. The PC will not be in the front line of future mass communication devices.

Not only is the PC not appropriate for viewing all data, neither is it appropriate that all digital data coming into the home should be consumed in its digital format. For instance, personalized newspapers could be delivered to an IP-connected printer in the early morning and printed out for each family member, to be read by them at the breakfast table or on their way to work or school.

(See media critic Vin Crosbie's recent OJR article on an intermediate step in this direction, the mass customization of editions relevant to each reader's interests, "What Newspapers and Their Web Sites Must Do to Survive.")

In publishing we are moving from a "print and distribute" to a "distribute and print" model, in which the "printing" is to a digital display. In between, as with all technology transitions, we will have less glamorous digital-analog hybrids such as the "distribute and print and distribute" model pioneered by satellite printing. (Of course, newspaper publishing has long had an analog-to-digital conversion as production moved to Atex and SII, and then to Quark XPress and Adobe InDesign, but the final output was still ink on paper.)

Limits of adopting pre-digital models

Even the PC/Web-browser-based news delivery model is ill-considered. The use of digital technologies and the Internet for mass communication has largely built on existing models for communication, or adopted and stuck to models from the early days of the Web. Not only are print and broadcast models applied to the Web, but the same formats are applied across platforms, including portable devices and interactive television.

This has been taken to extremes with the digital editions that have recently proliferated, including The Wall Street Journal's Today's Print Edition, and the UK's Digital Guardian, which Vin Crosbie rightly characterizes as simple, unmalleable facsimiles, lacking digital enhancement.

Part of the attraction of digital facsimile editions is that they follow a layout and navigation system that has been well-established for decades. None of the Web-based editions have been able to pioneer a more powerful "interface." For example, a basic convention in print is to use larger type to emphasize the importance of a story. This convention has no established equivalent online, where lists dominate. And a story at the top of a list may simply be the most recently posted, rather than the newest. Structural and navigation models are further confused by the mixing of online "archives" of printed editions with online-only content.

Digital facsimiles that build on all the well-known properties and behaviors of printed publications could be valuable if they did this well and added some extra "magic" by exploiting the available computational power and network. But as they are, the future of online news won't be digital facsimiles, not least because a portrait -- or vertical -- format is hard to navigate using landscape -- the dominant or horizontal -- displays. Newspaper formats were shaped by their medium, as future formats will be. The Digital Guardian, for example, is thoughtfully and elegantly designed -- it just doesn't solve a real problem or do anything very well.

(See Guardian Unlimited Editor in Chief Emily Bell's explanation of the Digital Edition)

Of course, there will be transitional stages, and it is quite possible that in the future a significant readership will access digital facsimiles on tablet PC-type devices. And digital versions are ideal for publishing models in which editions are distributed for printing at local sites, such as at hotels, speeding delivery to regional and foreign readers and allowing expatriates and travellers to keep up with home news.

Ironically, some print publications are actually doing a better job than most online publications of solving some real problems we are facing. One current issue is how to help readers swamped by editorial content to understand at some level stories they skim, without their having to read the full text. Here, BusinessWeek in print excels, with a rich contents page and box-out summaries within each story.

Ease of use is key

Beyond these fallacies, the technology to create compelling new models for mass communication already exists. But too much focus has been directed at the technology and not enough on understanding the more significant barriers to the creation and adoption of new products. New products need to be groundbreaking, but they should also be as easily understood as possible and as easy for people to use as they can be. They should also support the contexts and situations in which people are likely to use them.

They may be cooking in the kitchen, collapsed on the couch, relaxing and reading at the dining table, hurriedly hunting information in the rain or discreetly diverted in a meeting. What is the right device in each situation? Perhaps a surface-top screen, an interactive TV, a tablet PC, a cell phone or a networked PDA. The key to finding the best solution is to be familiar with your audiences.

This brings us back to human-computer interaction (HCI). Andrew Zolli, a New York-based forecaster and design strategist, observes in an article in the Association for Computing Machinery's journal Interactions entitled "Can HCI deliver on its promise?" that "the very thing that makes traditional mass communications platforms (such as television, radio, newspapers and magazines) so successful is how completely they have solved their respective interface challenges." This achievement is underappreciated by mass communicators and is a key area in which people with human-computer interaction and design skills can help these new industries develop.

The mass communication industries ignore a number of basic HCI principles. Beyond ease-of-use is the need for interaction design that works well and is satisfying for the user. This concept of satisfaction is a key aspect of almost every mature product or service and recognizes that our experiences are not just functionally based.

Not only do mass communicators largely fail to deliver ease-of-use to their audiences, they also fail to deliver it to their own editors, which affects their productivity and job satisfaction. Being able to view and author in the same environment and directly manipulate material are important ease-of-use characteristics of online editing tools. This concept was integral to the first Web browsers.

Weblogging tools, with their "edit this page" model, have resurrected this approach -- and exposed the awkwardness of more expensive commercial and content publishing tools. Weblogging tools, however, are poor for manipulating complex material, workgroup editing and workflow. These are strengths of products such as Macromedia Dreamweaver and Adobe GoLive, though they lack the ability to flip a page from view to edit and the flexibility of working from any Web browser. Slowly, these attributes of publishing tools are coalescing.

Context of use is critical

New media publishers should not automatically assume that news consumers will be at an online PC or interactive television for long hours -- even if they were their preferred devices. People are mobile, both around the home and out of the house. Their varied activities are not focused on one location or device, and their choices in each case will be determined by desires for privacy or company, supporting artifacts they need for an activity, the characteristics of a device and competition for use.

This issue has been addressed in limited ways, most prevalently by the provision of "printable versions" of stories, which at least recognizes that screens are uncomfortable for reading and that people like to read in more reflective modes and locations. Products that turn PDAs and cell phones into portable reading devices, such as AvantGo, Roundpoint, Weblog readers including Hand/RSS, and mobile WAP and Web browsers indicate another way forward. Text-messaged alerts are another. And audio summaries of newspapers, such as's New York Times digest, which can be listened to on an Apple iPod, are another still.

As well as appreciating and designing for particular contexts, mass communicators need to consider how to support people moving between contexts or formats. The most common scenario, the print reader who wants to send a friend or colleague a link to an interesting story, is -- save for Wired magazine's long abandoned CueCat experiment -- entirely unsupported. Instead the reader must note some unique keywords from the story, go to the publication Web site and use the keywords to search for and, if they are lucky, locate the story (or browse to find it if they feel brave). To their credit many publications at least support the next step with "e-mail a friend" features -- though accurately entering one's friend's e-mail address is another matter.


The digital mass communication industries need to take a long hard look at their strategies, which are often informed by technology-driven fallacies, or conservative interpretations of new media.

In these larger challenges they have a lot to learn from HCI and design. These disciplines can inform future publishing models.

In previous eras, this learning has been acquired in a Darwinian fashion. Ambitious players should be considering how to circumvent evolution and push forward the intelligent exploitation of new media forms.

Also see sidebar: "HCI Gains Visibility as Digital Media Mature"

Editor's note: This is the first of a series of articles by Macdonald on human-computer interaction and design issues and how they relate to online journalism.

Related Links
Association for Computing Machinery
Digital Guardian
Guardian Unlimited: Emily Bell's explanation of the Digital Edition
Interactions (ACM journal)
OJR: "What Newspapers and Their Web Sites Must Do to Survive"
Sidebar: "HCI Gains Visibility as Digital Media Mature"
The Wall Street Journal