CHAPTER 2

Usability and how to achieve it

Usable information design certainly incorporates how most people perceive and comprehend. But it's best known for taking audience concern to a deeper level—examining how each particular audience varies from others and within itself—and designing accordingly.

An understanding of usability, and how you achieve it in your designs, begins with its multilayered definition. The word itself—dissected into use ability—provides a clue, but leaves out the essential parts: the people and their experience. Usability is the ability of an object or system to be used with satisfaction by the people in the environment and context the object or system is intended for.

Let's focus on satisfaction for a moment. Being able to fill out an insurance form or get an online boarding pass doesn't mean the experience was satisfying. By designing an experience for your audience members that's as satisfying as it is functional, you'll improve their day and probably earn their loyalty.

Usability combines useful content with a presentation and format that lets the audience easily, even intuitively, navigate and understand it. And although you might hear the term usability used most commonly with Web sites, it's hardly limited to them. Consider the audience's experience in every project you design.

To ensure your projects are usable, strive to design information that encompasses what Whitney Quesenbery of Whitney Interactive Design refers to as the “five E's of usability.” Make your design:

1. effective. Can audience members—users—complete their task?
2. efficient. Can they complete the task with a reasonable amount of time and effort?
3. engaging. Is working on the task satisfying, maybe even pleasant?
4. error tolerant. Does the design help users avoid and recover from mistakes?
5. easy to learn. Is the design based on what users (see the sidebar to the left) already know? And are elements used consistently to reinforce what users learned when starting the task?

Keep the list in mind when you encounter, design, or research any task-oriented visual information. Why research? Usability rarely happens on its own. To know what users will find effective, efficient, engaging, error tolerant, and easy to learn, it helps a lot to know the users, their environment, and their context (abilities, background, knowledge, habits, and preferences), then design with that knowledge.

Research basics: Getting to the heart of the audience

This chapter—and the book—includes case studies that used a wide range of techniques you might use to find out about your audience. They fall into these major categories:

• Find out all you can from your clients:
  • Interview the people who give you the project.
  • Talk to people who have regular direct contact with the audience in person, by phone, or online. That might include customer service or sales reps, clerks, registrars, bookkeepers, and so on. Ask your clients to arrange access to them.
  • Look at the clients' logs of the audience's phone calls, e-mails, Web site paths, and search engine entries. (If the clients aren't keeping and using these logs, urge them to start. You might help them to design a system for capturing this valuable intelligence.)
• Do or commission your own research. For example:
  • Interview audience members in person (one-on-one or in a focus group), by phone, or online.
  • Conduct a survey (in print or online).
  • Observe audience members working with your materials and those of other companies.
• Look at outside research about the audience.

Interview your clients and their colleagues

Whenever you get a design project, always ask your clients for a detailed description of the audience members and their needs. Here are some essential questions to start with:

• Why are these people in a position to interact with the design? What need, problem, or question brought them to it? What's their task? Even if the answers seem obvious to you before you ask for them, they might show you a flaw in your assumptions early on.
• What else do audience members see on the topic from other sources?
• What else do they see from the client? (Look at the proposed project in the context of other materials to see whether it would make sense to combine, redesign, or eliminate some.)
• What does the audience already know?
• What problems or challenges does the audience have?

Educate your clients

Clients who haven't looked at the audience might not be able to answer all of those questions. Or you might have to explain why you're even asking. If so, consider it worth the effort to train your clients in the value of research and serving the audience members' needs, not just the clients'. Consider it a long-term process; you can use this book and its case studies as support.

You also might need to educate your own team members, suggests Sanjay Koyani, a senior usability engineer at the U.S. Department of Health and Human Services. Teach them not just about the need for testing but also its goal: “making sure we have a deep understanding of the users, their priorities, and our measures of success,” rather than just that the site works on the servers.

When designer Paul Mijksenaar recommends that clients test their designs with the audience, “most of the clients say, ‘Well, you are the expert, you should know’” what's right without research, he says. “We can say if the details are too complex,” he notes, but not “how people behave or the meaning of the sign … we can't foresee how people will react to a new element.”

What to learn about your audience and why

Learning the audience starts with knowing their demographic categories, such as age, gender, education, income, and marital status—information that's relatively easy to get secondhand. But it also means delving deeper into the audience's psychographics, such as attitudes, ambitions, habits, preferences, skills, and details of daily life. Psychographic info is tougher to get at, but it's worth the effort.

Knowing how your audience members think, act, and feel will tell how they're alike and different from each other and from other audiences. It will:

• keep the focus on what the audience needs instead of on what the client wants to tell them. (And any utterance of “What do we want to tell them?” should always trigger the better question “Why should they care?”)
• serve as a benchmark for every content and design decision
• help to eliminate or minimize the effects of different agendas within the team that can derail a project
• help you and your colleagues think about and speak to the audience as individuals (instead of as “you guys,” as an insurance rep's voice-mail message addressed callers: “If you guys are calling for claims forms …,” which seems to reflect the rep's lack of concern for her audience)

But words alone won't convince the audience you care. No one's fooled by a utility company's assurance that “your call is important to us,” especially when it repeats during a long wait on hold. Although this example is not visual, it falls under information design because it's part of the audience's information-seeking experience. Consider the whole picture to give your audience a consistently positive one.

Research will also help you:

• anticipate what will work for the audience members
• avoid alienating them
• develop empathy for them
• connect with them
• do your part to make life easier for the people you serve

How to learn about your audience from your audience

Research techniques include:

• interviewing the audience
  • one-on-one or in groups
  • in person or on the phone
  • print or online surveys
• testing the audience
  • Create project-related tasks for audience members.
  • Observe them doing the task.
  • Ask them to tell you what they're doing as they do it.
  • Interview them after the task about their experience.
  • Test before you design—perhaps using an existing similar project—and during and after design.

For a step-by-step walk through testing for one project, see—and do—the swim schedule project on pages 42–60.

Testing tips … and limits

Tests don't have to be expensive or complicated. In fact, it's better to keep them low-cost and simple to encourage more testing, according to Don't Make Me Think by Steve Krug, who focuses on Web testing. Ask people to do typical tasks you set for them, or, he suggests, you might let them set their own tasks. You also can show people rough sketches and ask for reactions.

Krug notes that there's a limit to what testing will do for your project. It won't tell you what everyone wants, because there is no “average Web user.” Krug's observation of test participants over the years confirmed that individual Web use behavior is both unique and idiosyncratic. Good design allows for such diverse behavior.

According to Krug's commonsense approach to testing:

• Test early because it's harder to change things at the end. Before you design anything, test with sites of organizations that are similar to yours.
• Test often, even if it means using participants who don't represent the audience in some ways for some of the tests. For example, if you're designing a Web site for realtors, you could test navigation throughout the design process with nonrealtors who have Web skills that are similar to your audience's in between tests with the actual audience. (Janice Redish also believes in frequent tests, even with only a few people, but only with people who represent your key audiences. But, she says, “most Web sites are meant for the public, and it's not too hard to find the public.”)
• Test inexpensively. Testing can be informal; it doesn't need a lab or scientists.
• Test after making changes.
• Videotape tests.

Interviewing Guidelines: What to Ask and How

The typical survey question is a variation on “What content [articles, sections, answers] do you want?” Although you might get a few useful answers from such a direct question, don't be surprised if you don't get more. In fact, such a question at the start of a survey could suppress overall responses. That's because audiences often don't know what they want until they see it. They expect you to know what they want; if you don't provide it, expect them to look somewhere else.

Find out what the audience wants by reaching their heart: Ask psychographic, less-direct questions and observe the audience. The most valuable questions are oriented to emotional experience, such as “What frustrates you most?” or “What is your greatest challenge?” Compare that with the less effective wording of a similar question that engages the logical mind first, ‘Tell me the five tasks you most would like to replace with a robot,’” says Nathan Shedroff, information designer and author of Experience Design.

For one subscription newsletter, responses to the challenge question painted a vivid picture of the audience that pervaded every design and editorial decision. Although subscribers were busy, they were so pleased to be asked that they responded in unprecedented numbers. They also responded from the heart and in depth, exceeding the provided space to write pages of their own stationery. Those letters translated into dozens of articles that targeted and solved subscribers’ expressed problems.

Ask other questions to delve into audience members’ motivations (see sidebar to the right), goals, and reward systems. But tailor the questions’ choice and wording to the audience's tolerance level. Depending on the audience and the situation (as always), someone might be more likely to answer personal questions one on one than in a focus group or on the Web. Or not. Find out.

Find out what (else) audience members read, what Web sites they visit, where they shop, and how they spend spare time. But asking questions isn't the only way of getting at psychographic information. Shedroff likes to create a game and watch how people play it, so he can gauge “their problem-solving skills … or their emotional state through the process.”

Observing the audience: What else to find out

Burkey Belser, partner of Greenfield Belser, Ltd., a design firm., combines university-based observational research with his firm's own research to find out how people interact with information. He looks at such essentials as:

• How does a consumer read a brochure?
• How does a consumer read an ad?
• How much time do you have before the reader loses interest?
• What part does the reader look at first?
• What can you learn from tracking eye movement?
• How do the human body and mind react to stimuli?
• How do psychographics and demographics influence consumer behavior?

How does a consumer read a brochure? From back to front is the typical pattern, Belser says. That means “if you anticipate telling your story in some strictly linear way,” you might lose the readers before you can engage them. He cites magazines such as the New Yorker, Life, and Playboy (“… the skin magazines are actually pioneers in many ways in information design”) that engage readers with full-page pictures in the back. By contrast, most newsletters put “the best stuff in the front and the worst stuff in the back. You really need to create a rhythm and a movement for the reader throughout the piece.” (See the sidebar on the facing page.)

How much time do you have before readers disappear? You have “maybe five seconds to capture readers and encourage them to … stop or slow down or do anything but go back to the scanning mode,” Belser says. Every page must have a goal and “something for (the readers) to grab hold of.” To help clients understand the concept, show them, he says: Hold up a page for five to ten seconds, then ask, “What was your take-away message?”

For example, the intended take-away message of the brochure to the right is that the advertiser can help law firms with an important goal: recruiting recent law school graduates. The brochure demonstrates research that found how students really choose their employers (nearby location and a salary high enough to let them repay student loans quickly), versus how law firms might think they do (impressive clients and a quick partner track).

Some researchers go even further than observing the audience, to try to see through the eyes of the audience. They give audience members disposable cameras to record their surroundings without directing their process.

The cover of a design firm's brochure passes the five-second take-away message test. Design firm: Greenfield/Belser Ltd. Designer: Siobhan Davis. Creative director/Copywriter: Burkey Belser. Production artist/Illustrator: Brett Mannes. Photographers: John Burwell, Jason Hendricks.

CASE STUDY: Drug label design relied on research at the mall

Confusing information design can frustrate people. Confusing labels on drug packages risk much more. Such labels have to be clear to the broadest possible audience, including people who have low literacy levels and/or poor vision. So it's good that the design of the “Drug Facts” label came after that of “Nutrition Facts.” For both Food and Drug Administration projects, research included “mall intercepts” (interviewing people at shopping malls about their reaction to presented designs). But the nutrition label test participants saw only a binder of big printed-out designs, Belser says. “An 8½-by-11-inch label in a binder that you stop me for is going to get a very different read than a tuna can that I'm running down the aisle for.” So for the drug label research, he put actual-size labels on typical products such as allergy medicine. The more realistic format made a huge difference in the quality of response, he says.

Adding to the project's challenge: Unlike nutrition labels, which need little more than calorie, fat, and vitamin counts, drug labels have to include “indications, contraindications, warnings, contents … probably three times as much to say in half the space,” Belser says. So the type had to reduce accordingly. “The best we could do was 6-point,” compared with 8-point for the nutrition labels. Even with 6-point type, he says, on some products the designers had to extend the available space with hang tags.

Although the research for the first labeling project informed that of the second one, the agency required a different look for the two projects, despite the success and familiarity of the first one. Half of the process “of getting people to understand labeling is when they see it again and again,” Belser says. “That learning had already occurred in America [with nutrition labeling], so why not just copy it down to the last detail?” He believes the consumer “would benefit from a nationwide labeling standard” on everything, including “mattress tags.”

According to FDA's Office of Nonprescription Products, FDA developed and tested different labeling formats through industry and consumer groups. To adhere to the requirement, Belser focused on the features of an informative label: strong typographical hierarchy, legibility, and contrast, which is especially important without color to help distinguish the information. He paired a 14-point black with a 6-point medium. Lines that separate sections also add contrast, but not too much: A thicker line would steal space from essential type. The lines around the drug info are essential to “prevent the [drug's] manufacturer from invading that territory” with marketing copy. The designers planned drug information labels to fit all possible package sizes and shapes.

Six-point type is enough to give you a headache on labels for U.S. drug packaging, but it's bigger and clearer than on earlier labels. Required legal and scientific language along with the contents limit the size. Bigger, bolder, and italicized headings provide contrast—the title's in 14-point type, subtitles are in 8-point. The 2½-point dividing lines separate sections from each other and the label from the package, without stealing text space. Design firm: Greenfield/Belser Ltd. Client: Food and Drug Administration. Creative director, Designer: Burkey Belser.

CASE STUDY: Finding the tasks in a Web site

Web sites used by both an organization's insiders and an external audience tend to grow well beyond their value. That can happen because every department wants a prominent Web “presence.”* Or a department wants to show everything it does. And maybe employees want to store everything there so they can easily find materials. So they add sections and pages without a structural plan to govern those additions.

The result is a site whose structure can't support all of its content and whose useful content hides from frustrated audience members. In fact, even a site of a reasonable size might contain “hundreds of different tasks people can do,” says Sanjay Koyani, a senior usability engineer at the U.S. Department of Health and Human Services.

To find out which tasks and needs to emphasize in the redesign of his agency's Web site, Koyani's team created a process that involves users “through market research, interviews, and discussions, by looking at terms they enter into our search engines, through Web analysis” (what they click on during actual Web visits), and usability tests. The team strives for “a continued understanding of how these folks are interacting with the information, what's important to them,” and whether the site continues to work for them.

The audience's searches and survey responses determined the order of the topic categories on this home page. The designers put what's most important to the audience on top. United States Department of Health and Human Services. Design team leader: Sanjay Koyani.

One of the specific insights that testing generated was the value of templates. So the team used pages that tested well as models for new pages. That way, users “don't have to relearn all the navigational stuff on each new page.”

Nor did testing end with the design. Koyani advises developing a long-term plan for checking how people are doing with the site, and identifying new needs. To accommodate those new needs, his team designed a features section. “What you never want to do is just sit there and say, ‘We've got a good site—people will find it if they want to.’” Instead, he says, make sure people are having a great experience.

CASE STUDY: Task-based focus groups help to redesign car manuals

A senior official at Ford couldn't figure out the clock-setting instructions in his car's owner's guide. He decided that if he couldn't understand them, at least part of the audience couldn't either. That led to testing, rewriting, and redesigning the manual and other customer materials at the former Document Design Center (DDC) in Washington, D.C.

Before and after the redesign, users were asked to use the manual to answer questions and do tasks. The tests pointed out problems that needed correcting in the redesign. The tests also showed that people aren't alike in how we look for information. Some begin with the index, some look at the table of contents, and others plunge right in, says Carolyn Boccella Bagin, who worked on the project at the DDC. The third group would appreciate tabs.

Other research-inspired improvements included:

• changing the format from horizontal (wider than tall) to vertical to make it easier to hold
• rewriting the information in plain language
• reorganizing it into digestible chunks
• adding categories to the table of contents based on users’ needs
• adding white space
• indenting warnings and adding a vertical line to make them obvious

CASE STUDY: Surveys and recipe tests lead to a usable cookbook

“How do you cook? How do you choose a cookbook? What do you think of this recipe?” Before they designed a cookbook, Maria Giudice and Lynne Stile surveyed home cooks and asked them to test recipes. The answers, plus looking at outside research and other cookbooks and recipes, guided the designers to:

• list ingredients in the order of use
• group steps the way cooks would
• put ingredients in a different color than the steps
• group recipes by the amount of time they take
• put substitutions in the margins

Researching the audience “not only tests … or validates your assumptions, it allows you to come up with new things,” Giudice says. For example, when the designers asked audience members to test recipes, feedback split regionally. The East Coast wanted “more butter, more cream,” the West Coast wanted less of each, and the Midwest wanted more meat. Home cooks who live on the East Coast (where the cookbook author's restaurant was located) were the book's primary audience, so their preferences took priority. The substitutions listed in the margin were for everyone else. But the designers didn't label the substitutions by the region of the people who needed them. Instead of saying, “If you're from the West Coast, this is for you,” they'd say “Instead of cream, you can use yogurt.” That's a deliberate distinction, Giudice says. In information design, you shouldn't “tell people who they are” or make potentially insulting or inaccurate assumptions.

CASE STUDY: Universal how-to graphics, minimum of type

A company needed clear, keepable, accessible instructions for operating its products—handheld or wearable scanning devices that ring up purchases, check prices, or take inventory. The goals for the instructions were to keep:

• users from injury, including back or repetitive-stress problems
• users’ employers and the manufacturer from liability and added costs resulting from unsafe usage
• the products working properly

The main audience—the people who use the devices—vary widely in native language and in education, economic, and occupational levels. They range from inexperienced seasonal employees to “power users who … have figured out the shortcuts,” says Sudhir Bhatia, senior manager of design research in the Industrial Design and Human Interface group at Symbol Technologies, Inc., which makes the devices.

The client's research looked at the “entire experience when someone gets a product.” It found that a previous attempt to promote safety—instruction books that shipped with the devices—didn't work. The order departments often threw out the books with the box. Even books that survived failed to do their job, because they hid the essentials in too much information.

So the problem became, in Bhatia's words, “How can diverse and frequently changing populations of operators learn to use these devices effectively and safely?” Asking the question, the solution seemed obvious: “Use clear, accessible, and keepable instructions that work for every level of user.” It also triggered another set of questions to address with research and design.

The company commissioned a poster for each device to display in a facility's break room. One side of each poster shows the best angles and body postures for safe scanning. On the other side, a “Quick Start Guide” answers another worthy question: “What does the user need to get off the ground?” It contains the most relevant content from the book, such as instructions for changing or charging a battery.

A pair of posters shipped with the device lets the recipient display both sides if they want to (see page 70). But it's the safety side that's most important (page 71). Most users sign scanners out from an IT (information technology) office each shift, so they don't have to deal with the set-up or maintenance, Sylvester says. And “the processes are simple and easy to remember—batteries only fit one way and are easily removed; scanners are set up once for their task and software type, not reprogrammed daily.”

The value of nonverbal communication

The posters emphasize graphics over words, for at least three reasons. Graphics:

• project more clearly than words in break rooms with dim lighting or too much distance between chairs and walls.
• communicate across languages. In the posters, type is limited to the bare essentials to make instructions perform nonverbally as much as possible.
• can precisely describe objects. They show where to put the battery, while words can merely tell.

Depicting the products

The designer chose illustrations over photos for:

• clarity. Some of the emphasized details don't show up as well in a photo as they do in a black-and-white illustration, says art director/designer Mark D. Sylvester, Interrobang Design Collaborative. In part, this is because the products tend to be “on the dark side … to keep them from showing dirt, scuffs and scratches.”
• emphasis. Although you can emphasize a feature on a photo with an arrow or a circle, in an illustration you add even more emphasis by increasing the feature's size or by simplifying its surroundings. (With a program such as Photoshop you can do that in photos as well, but ethical issues come up, such as “Has the retoucher misled the viewer?” Users might wonder why their device looks different from the one in the photo.)
• time. “The product often is still being designed while we're getting these projects done,” Sylvester says, so prototypes might not be available in time to incorporate photos of them.

A manufacturer found out that its instructions on how to quickly and safely use its portable scanners weren't getting to the scanner operators. So a series of instructional posters, designed to function across languages and to be displayed in work and break areas, replaced bulky instruction books. On one side of one product's poster, the “Quick Start Guide” shows how to insert and charge a battery and install accessories. (On these posters, the second color is actually red, not blue.) Design firm: Interrobang Design Collaborative. Client: Symbol Technologies. Art director/Designer: Mark D. Sylvester. Illustrator: Richard L. LaRocco, Graymatter Illustration

The other side, the “Recommended Usage Guide,” demonstrates the body postures that will prevent repetitive-use injuries. Design firm: Interrobang Design Collaborative. Client: Symbol Technologies. Art director/Designer: Mark D. Sylvester. Illustrator: Richard L. LaRocco, Graymatter Illustration.

Depicting the user

When materials show a person using a product, what should the person look like? Art directors often struggle with decisions about the look, gender, race, and age of both live and illustrated models used to convey a message. This device's instructions use an illustrated man who looks racially ambiguous, which might help to avoid a racial stereotype about the products’ users, and to connect with a range of users.

Of course, this might be true of only male users—the absence of a female demonstrator might suggest to some that the manufacturer thinks women aren't qualified to do the job. The designer could avoid that suggestion by using a woman in some illustrations. But that decision opens up more cans of worms: Which pictures? Who would demonstrate the do's versus the don'ts? And who would be shown bigger, who smaller? You also can argue that, like nonparallel construction, a different demonstrator might draw the user's attention away from the action (where you want it) to the person (where you don't).

Designing to enhance the user's experience

The poster arrives folded, and even when it's flattened for hanging, those fold marks remain to imply a grid, Sylvester says. He designed within that grid in case people refolded the paper “if they need to, and a lot of times they do.”

Gray backgrounds help to separate different categories of information on the page. They also help to minimize contrast on secondary information. As a background, they create contrast from which black-outlined white objects “pop.” That's another concession to the dimly lighted break rooms Bhatia and his colleagues saw in client visits. “Some of these back rooms … have one bare lightbulb hanging over the work area. It's not optimum lighting,” Sylvester says they told him. So they're wisely “conservative about how dark a gray can be behind a field of copy. A 50-percent gray background with black 14-point text is not contrasty enough.”

The designers didn't worry about matching the gray tones from product poster to product poster. “If somebody says, ‘Gee, they went light to dark on this poster and dark to light on this one,’” Sylvester says, “it doesn't compromise the information.” But red is a consistent element throughout the poster series as an emphasis and warning color, as well as a reinforcement of the brand identity. The backgrounds of bar codes also match on every poster. They have to be white for scanners, which demand higher contrast than even human eyes do.

Sylvester chose Univers type for its versatility. He underlined some headings, but no more than four words at a time because underlining poses a a legibility challenge. “The g's do get closed off a little bit, but I think you tend to read through it.”

What the people in the break room think

The lighting detail wasn't the only intelligence gleaned during research visits to the field. Answers to the industrial designers’ questions informed the poster's content and design—“things like ‘What kind of drop rate do you want when you're on the forklift and you're up on the third level of the storage area, restocking the bottom area?’ ‘What's working for you?’ ‘What's not working for you?’ ‘What would be more comfortable?’” Sylvester says.

Model numbers versus words

Why use letters and numbers instead of words to name products? It's tough to recall and distinguish model names such as LS/VS 3408 and MC9000-S. Indeed, Bhatia says, “we've gone back and forth on names,” but the company standardized on numbers because “you can create a hierarchy” with them.

CASE STUDY: Dealing with conflicting audiences’ needs

You can't please everyone, you've seen in this chapter, so make sure you're pleasing your primary audience. But when audiences’ needs conflict, sometimes those of a secondary audience take over. That's what happened in this case:

Tests by Discovery Channel TV showed viewers clicking away when programs’ closing credits rolled, a company rep says. So the network tested removing the credits. That worked for the viewers, but not for another important constituency—the network's outside program producers. For them, the network proposed alternatives, such as a brief closing screen that directed any interested viewers to the Web site. There the network would post each program's credits for six months after the program's airing.

But producers and industry groups still complained, so on-air credits remain. The network continues to experiment with alternatives that work for everyone, including opening instead of closing credits.

This project from the same design firm as the previous image relies even more on pictures and numbers to cross cultures without translation. In twelve steps, a folded manual shows how to adjust the snowboard bindings it comes attached to. Red (where the blue is here) stands out on the otherwise black-and-white piece to show action and direction, and to emphasize various steps and parts. Design firm: Interrobang Design Collaborative. Client: Burton Snowboard Company. Art director/Designer: Lisa Sylvester. Illustrator: Richard L. LaRocco, Graymatter Illustration.

CASE STUDY: Keeping the audience in touch

When recycled paper became popular in the 1990s, the editor of a newsletter for graphic designers knew from research that the audience needed to know the paper's pros and cons. So the editor not only wrote about the paper but also let the audience test it. She printed five consecutive issues of the publication on a different variety of recycled paper each time. (See the images on the following two pages.)

Getting inside your audience through personas

Use your psychographic research to group audience members into categories. Then you might find it useful to create what usability people call “personas” to represent your primary and secondary audiences. A persona is an imaginary person, with a name and a life story, whose traits represent a typical audience. It's used to help everyone on a design or development team speak to the audience's needs.

Build personas from individuals’ words, attitudes, even appearances to give audience members a real voice in the project and its development. Use them to form a picture of your audience members that's so accurate and convincing you think and speak of them as one person. You refer not to “them” but to “he” or “she,” or, better, by the name you've given him or her. When you talk to them, you use “you.” You might even find a picture (maybe from a newspaper or magazine) of the person who represents your primary audience. Display it where you work, and distribute copies to your colleagues.

Buyers of printing paper got to sample five makes of paper in as many issues of a trade publication. In the issue following the experiment, an article, including a pictorial guide and a table helped readers compare the five. (See details on the following page.)

The process of identifying and building personas, shared by everyone in the project team, can:

• add dimension, reality, and familiarity to major audience types; add recognizable human face and characters
• solidify the shift in focus from the client's needs to the audience's needs
• redirect the source of design and content decisions to the audience

If you use personas, go beyond the surface, information designer and author Nathan Shedroff advises. Personas are no more valuable than the research that goes into them. He cites typical—and useless—personas in Web site development as “‘Mark is twenty-eight years old and he's a software programmer. He's young and energetic and curious. He loves computers and he plays a video game.’ We create this whole fictional story about him, and move on to Heather, ‘thirty-six and a writer who works out of her house.’”

Here's a closer look at the pictorial guide.

The problem is, Shedroff says, although the demographics might shift in the stories created, the psychographics don't. “They all end up being the same person,” the minority who are in love with technology, “curious, not afraid of the computer, and who love their job and their work.” What's lacking in these cases are realistic, insightful profiles, such as people who are mean or who just don't understand. For them, realistic scenarios will “take people through the help system, or have people call up customer service” because they can't figure something out.

And here's a closer look at the table. How would you improve the table? (Hint: Eliminate unneeded repetition.) Publisher: United Communications Group. Newsletter: In House Graphics. Designer: Ronnie Lipton.

And although even made-up personas help a development team talk about Web site navigation, Shedroff says, authentic ones “highlight potential problems” and go beyond “Where do I click?” to “Would I do it this way?” The personas you develop “should be the synthesis of similar people and their problems and attributes … that come from real interviews that included psychographic questions.”

You begin to see audience members you might not have considered before, Shedroff says, such as those who have trouble using computers. “They don't hunt around the screen, and when they can't find the next thing to do, they go into panic mode … or they turn off the computer.” When you understand how they approach a problem, you can create a meaningful test scenario, he says.

Personas personified: an example

Let's take a look at how personas work. Meet Frances, a single twenty-two-year-old woman about to buy her first new car. You're doing a Web site for a car dealership that finds her typical of the audience it wants to pursue. So you'll want to find out as much as you can about Frances. And you'll want to see how she's like and different from others who seem to be in the same audience.

To create the persona, you'll interview and observe Frances and other audience members, and interview management and staff members. But before you begin your research, notice and eliminate any assumptions you might have about Frances, so you'll be open to hearing the real story. For example, maybe you assumed she wants a car that's:

• cute. (Uh-uh. When asked, she describes a homely and reliable car that won't attract carjackers in the city or police attention on the highway.)
• compact. (No again. Her car should be big enough to take along three friends, but small enough to save on gas. Here she starts getting passionate about the world's energy problems and briefly considers getting a bike instead. Now you're getting at attitude; this might be the time to probe gently for any contradictions. Maybe being seen as environmentally conscious is more important to her than actually being that way if it means giving up what she wants more.)
• affordable. (Most people want that, but what's considered affordable varies from low-range to expensive. And not only will she haggle, but she also enjoys haggling.)
• low-maintenance. (Reasonably; she'll do oil changes.)

Other questions and observations will help you get deeper inside Frances's head and heart:

• What else does she like?
• What does she value?
• What are her goals?
• What does she read?
• How does she dress?
• Where does she live?
• What are her image and expectations of the car-buying process?
• What does a car represent to her?
• What's her history/experience with cars, and car shopping and buying?
• What have been her biggest challenges with cars, and car shopping and buying?
• How often and when will she use the car, and to go where and how far?
• What are her information-seeking habits and skills?

More: Does Frances have a specific car make and model in mind? If so, find out where she first noticed it and what she likes about it. And how does she shop? If she says she does at least part of her shopping on the Web, how does she search? And what questions does she want answered, in what order?

Part of avoiding assumptions is favoring open-ended questions over multiple-choice ones. For example, ask, “How do you shop?” not “When you shop, do you drive to dealerships, talk to friends, look in the newspaper, go on the Web, use more than one method?” That's because multiple-choice questions limit the range of possible responses; the real answer might be something you never thought of. Also, hearing respondents’ own words will give you greater insight into their thought patterns.

And observe, rather than just interview. Observing Frances's and others’ search process will tell you even more than asking about it. That's because we're not always aware enough of our behavior to describe it accurately. Both methods will help you develop a clear picture of your typical audience members that will help you design well for them.

CASE STUDY: More about how to use personas

To create a health and nutrition book that would attract and serve its audiences, its information designers developed three personas. In any project, there might be as many as five, because most products, including publications, have more than one audience. After you identify those audiences, also identify—and focus on—the primary audience.

“I think people make a real mistake trying to be all things for all people, because you're going to be half-assed for everybody,” says Maria Giudice, principal of HotStudio, the design firm that worked on the nutrition book. “If you satisfy this person who represents the majority of the audience 100 percent of the time, the others might be satisfied 50 percent of the time. You have to allow them to be successful as well, but you're not dividing it up equally.”

Typically and optimally, those personas would come from direct qualitative audience research (interviews with the audience, one-on-one or in small groups).* This sort of research defines values and opinions and helps with identifying personas. For Web sites in particular, “we'll interview six to ten people the client identifies, we collect information from these interviews, and we turn them into personas,” Giudice says.

This project's budget didn't permit such research. So the designers could only interview the two authors (who are doctors) and marketing staff to identify the audience, product attributes, and business needs. They looked at competitive health books, magazines, and weight-loss programs. And they had to fall back on their own health and diet experience, and personal identification with the audience. That's better than no audience focus at all, but it risks false assumptions.

Testing the book with the audience would come only toward the end of the project, another budgetary effect Giudice would rather avoid: “When you test later, you don't get to make that many changes, so you better feel confident that your assumptions [were] right.”

Internal research guided development of the three personas:

1. Shelley, age thirty-four, married to Bert, two young children, lives in Minnesota. “She's fifteen pounds overweight … doesn't know why. She thinks she's got a slow metabolism. And it's hard for her to follow a diet because she's got kids. Shelley's husband, Bert, “is thirty-five and overweight, too, but doesn't seem to care about it.” (That's useful because the content can address the challenges and solutions when her family members don't share her concern.) Her activities: Tupperware parties, children's play groups (she might discuss the book with other Shelleys), Sunday school, reading (Family Circle and Parenting magazines), and TV (Trading Spaces and This Old House).
2. 2. Connie, a twenty-four-year-old single teacher, wants to lose five or ten pounds to wear a bikini without embarrassment. “Usually when people are a little overweight, there's a goal to weight loss: ‘I want to look better in a bikini or I have a wedding I have to go to.’” She drinks Coke and coffee for energy, but she feels too tired to exercise, which she feels guilty about.
3. Michael, a fifty-year-old book sales rep, will sell this book. He's well connected, overweight, and quietly concerned about his health. He's interested in the book, but he wants sound bites instead of having to read it. He wants to sell the book, but he also wants to know whether the book can help him.

Each persona has three or four quotes identified with it. Ideally, these would come from the words of interviewed potential audience members. But in this case, the designers had to make up the quotes (see the sidebar on this page).

For Shelley:

• “I've tried all kinds of diets and I can't lose weight.”
• “I exercise, but my metabolism is so slow.”
• “I want to see results.”

For Connie:

• “Why am I so tired?”
• “Is there something wrong with me?”
• “I'm hungry—what should I eat to bring my energy up?”
• “I'm stressed out.”

Michael has dual motivations:

• “I know which pills to take, but I'm not sure why I'm taking them.”
• “My joints always hurt.”
• “I want a source of information that speaks my language.”
• “The book helps me sell.”
• “Now I know what I'm talking about.”
• “Who wouldn't want to do this?”

Shelley's and Connie's needs proved so different from Michael's that the designers kept him out of the book. Instead they designed a separate sales and info piece for him.

Clear personas help the team evaluate content ideas through the eyes of the audience. They “help everybody focus,” Giudice says. Here, the team only had to ask, “What would Shelley want?” when proposed content went off track. For example, the designers knew she wouldn't care about such things as “glutamine … so there was a constant back-and-forth between what the users want versus what the doctors think the users want.” But that dialogue is useful for everyone. The designers listened carefully, because the clients “are the content experts, we're not.”

What also helped to keep the team on track was a diagram designers drew and taped to the studio wall. It read: “I feel tired, overweight, stressed out. Help me lose weight, gain energy, and relax. I'll need to pay attention to how I eat, exercise, sleep, breathe and relax. But … I'm busy, so I want something that's easy to understand, that's easy to follow, that (shows me) results …. So give me choices or a specific plan to follow.”

The two main personas inspired a list of questions whose answers turned into content for the book. Every spread had to answer one big question, such as “What is [the topic of the spread]?” or “How is it important?” The questions and conversations with the authors guided the content. For example, Giudice would say: “‘We're going to design a spread on ‘Why am I tired?’… Well, why am I tired? What's happening in my body?” The authors’ answers “became part of the book.”

What stayed out of the book were questions the personas would never ask. “It's more important to take out than to include,” Giudice says. “People tend to put so much stuff in that readers can't digest it.” (See Color Plates 7, 8, and 9.)

Burkey Belser uses research to avoid putting in too much stuff. When he finds out how much the audience already knows, he can see what to leave out. For example, in the law firm brochure, he avoided repeating info students will get from law firms’ Web sites. The printed materials only have to “communicate the spirit, the charm, and the working style of the firm.”

Keeping the audience in mind … literally

Some designers strive to put themselves in the audiences’ shoes even if or when they don't do research. Mental modeling and audience intuition are valuable skills to develop as supplements to, rather than substitutes for, actual audience research. For example:

• Joel Katz is guided by “modeling myself as different people, and different kinds of people.” He tries to “think of all the people who will see it and what their life experience has been.”
• Richard Saul Wurman tests clarity against two internalized audience models: his mother, “who wasn't really interested in everything,” and a “literate twelve-year-old boy” with a short attention span. Wurman also uses himself as the audience model because his projects come out of his interests. For example, he wanted to do a book about raising a dog because he couldn't find what he needed to raise his. “All the information was given out absolutely without the person—the owner of the dog—in mind … and nothing made any sense to me. Wow … what an interesting problem to solve: how to do a book on raising a puppy for the various me's in the world.” So his questions and their answers formed the content.
• Carolyn Bagin says she always begins by putting herself “in the place of a user, which I usually am.” Then she adds research, such as interviews, focus groups, and cognitive testing.
• Nathan Shedroff says a designer should have a “somewhat universal appreciation for people,” which isn't the same as liking all people or hanging out with average people. But he advises you to cultivate “at least a general understanding of what ‘average’ is in the culture you're designing for.” Designers are trained to do this in a visual way, Shedroff says. “They scan text and say, ‘This would make a great photograph or this would be a great cover.’ But they're sort of trained against saying ‘That doesn't make any sense’ or ‘This could be much clearer if I just did a diagram here.’” Understanding the audience often inspires going those extra steps.
• John Grimwade of Condé Nast Traveler magazine illustrates “for myself in many ways … it's not as if I know our readers personally. I'm doing what I think people want. It's hard otherwise to know where to begin.” But he also pays close attention to research, observing focus group participants’ reactions to magazine pages. He reads letters from readers who've used his guides to cities and museums. He also looks at how his audience members are different from him.

Caution: The audience is different, and you're not it

Avoid stopping at your own assumptions, especially the idea that you represent the audience. That's because as soon as you start to work on a project or for an organization, you start to know more about it and its language than its audience does. It's also because audiences are complicated: You might be similar in some ways, but never alike enough. And keeping the focus on yourself prevents you from looking at and responding to changes in the audiences and their needs.

Sanjay Koyani describes the risk when content and technology people make decisions that “match their internal perspective” instead of the audience's. These people tend to assume that if a design is intuitive to them, it'll be intuitive to the audience, he says. They also tend to assume that the audience has more skill and better equipment than it has. But if you test by bringing in “outside folks who have none of this institutional knowledge … they don't know what the devil we did.”

So for every information-design project, aspire to the Buddhist goal of “beginner's mind,” working to erase preconceived notions so you can experience a situation as if for the first time. Do so by talking to first-timers and current audience members, and being open to what you hear. That doesn't mean you need to ignore your instincts or any experience you might have as an adjunct audience member. They can be valuable at least as a starting point, but avoid letting them replace listening to the “unaffiliated” audience.

Expect resistance to change … inside and out

Often, listening to the audience in any form suggests the need for a redesign. Go ahead, but present the redesign only after you let the audience in on what you're doing and why. Here's why. (It'll be easier to understand if this has happened to you.)

As a subscriber, you feel possessive of your favorite magazine, so you're not happy when it shows up with a new look. “They say ‘new and improved’ on the cover and then you find everything you liked is gone,” says art director Ellie Barber about redesigned magazines she has subscribed to. Think of it as another usability issue: The magazine is no longer familiar, so you can't get around by habit anymore. It's as if someone redecorated your bedroom while you slept, so you trip over a chair on the way to the bathroom. Oh, you'll get used to it eventually, but only after the bruise goes away. It's also a perception issue—kind of insult added to injury—because readers often assume the change is not for their benefit (for example, that it's for cost savings that won't get passed on to them).

So avoid redesigning just because you're bored. When you work on a periodical, you see it every day, so it's no wonder you tire of it. But your readers probably see it much less often, and (again) find comfort and convenience in its sameness. Redesign only if the current design fails in some way:

• It bores or hinders your readers.
• It interferes with the content.
• It looks old-fashioned, suggesting out-of-date content.
• Its content category has changed.

You'll limit or even eliminate readers’ resistance to the change if you prepare them for it. If you can, even involve them in the change.

CASE STUDIES: Three ways to prepare the audience for a redesign

Build a bridge

Talk about change: A publication changed size, color, paper, frequency, editorial features, and even title in one issue. The journal of the Association of Governing Boards of Universities and Colleges went from 6 by 11 inches to 8½ by 11, from black-and-white uncoated cover stock to four-color coated stock, and from the title AGB Reports to Trusteeship, among other things. And it did all that without ticking off the readers. “Nary a soul complained about the switch,” says editor Dan Levin. (See Color Plates 10 and 11.)

The editor wrapped the new publication in a full-sized replica of the old one. Because of the old one's narrower width, it became a bound flap to show off one of the few common features of the old and new: the illustration. Under the flap, a letter from the president introduces and explains the new publication, its new editorial features, and its relationship to the previous, familiar publication. It's a verbal follow-up to an expressive nonverbal presentation.

Prepare the audience at every step

Here are other ways to introduce readers to a stranger in the mailbox. On the first redesigned issue, you might prominently show a picture of the old cover with a letter that details and explains the changes. Say that the publication changed to respond to readers’ requests, maybe to give them more of what they want in less time. You might back up your statements with results from your readership research.

Or put a notice on the cover of the new issue, as did Hemispheres, a magazine published for United Airlines passengers. (See Color Plate 12.) A circle in the lower left that color-coordinates with the cover art announces, “New Look! More Substance! More Style!” The president's letter inside gives details.

A still better way to link the old with the new: If you finish the new cover layout far enough in advance, show it with explanatory text in the final issue before the change.

Involve the audience at every step

The Bureau of Business Practice (BBP) took on the redesign of its thirty-three business-to-business newsletters in one year. The move was overdue; some of the designs hadn't changed in decades, and it showed. Before the redesign, each editor mailed a survey to each newsletter's readers. Beyond typical content questions, the survey asked readers what they thought of the newsletter's design and what they'd like to see.

You might not expect nondesigners to be able to answer such questions. But readers came through with useful results. Readers said they wanted:

• updated designs
• more charts and graphs to explain and break up text
• articles that are continuous rather than jumping multiple pages
• hole punching*
• less clutter
• shorter articles

Each newsletter's editor and designer incorporated the survey results into the redesign. And they prepared the audience for the change by sending two announcements with the redesigned issues: a “Welcome to [newsletter's name]’s New Look!” box on the front page and a letter.

The letter explains changes that sometimes involved a new name and focus, not just a new look, Barber says. It also gives this reassurance: “The information you've come to rely on is all here—but in a more attractive, easy-to-read format,” along with a list of features and their locations.

Survey 'em again

After you go to the effort of a massive redesign, you might be tempted to take a break. But it's time to check reactions, as BBP did. After readers got the redesigned newsletter, they also got a postcard with these questions, most about design:

1. What do you think of the redesign of our publication?
2. What do you like the most about our new appearance?
3. What do you like the least about our new appearance?
4. What additional information would you like to see in upcoming issues?

Responses to the first questions ranged from “eye-catching” to “you went too far,” with the majority mostly pleased with the redesigns. But complaints about distracting “fancy” effects caused the design team to scale back. They yanked screened photo backgrounds behind text, and anything that got in the way of finding story beginnings. Nor did the editors stop there in their efforts to show concern for the readers: after a few months, they checked in again.

More advice: For faster and easier-to-tabulate responses, do online surveys if you communicate with your audience on the Web. And make sure the survey device you send includes a photo of the new look.

The case against “art” versus content

You might wonder why the designers in the BBP example above needed to hear the scale-back message. When the previously restrained designers got a mandate to update designs, they sometimes tried out all their Photoshop tricks in one publication. That's easy to under-stand—but not to condone, as the audience reminded them.

Information design is graphic design combined with usability. It's the victory of:

• substance over sizzle
• function combined with form, not sacrificed to it
• how well information connects with the audience and delivers the content, over just how it looks
• building the layout from the content up, over just pouring in type
• respect for content and white space, over contempt for the first
• meaningful elements, over gratuitous decoration.

So work for the audience members on every project. Learn all you can about them, become an advocate for them, and encourage your clients or supervisor to do the same. Also read what you design. (Well, if it's a seven-hundred-page book on an unfamiliar technical topic, at least ask for a summary of the content and the audience.)

* You might point out to these people what the word presence means: just being there, not doing anything for the audience.

* Quantitative research, on the other hand, refers to info collected en masse, as in online or written surveys.

* When your readers ask for hole punching (known to offset printers as “drilling”) for a print publication, assume you're doing something right. The request says readers value your publication enough to save it in three-ring binders. By the same token, if you take it upon yourself to hole-punch your publication before you distribute it, the holes might send a subtle message that the publication's worth saving. (And send a free binder with the publication's name on it to store them.) On the other hand, you'll waste the touch and send a mixed message by punching holes in a publication whose info goes out of date the day after it arrives.