Information Architecture for Designers: Structuring Websites for Business Success

• Home
• What people say
• Table of contents
• Sample Chapter
• Templates
• Links

Buy now at Amazon.com, Amazon.co.uk, Amazon.ca, Amazon.de, Amazon.ja, order in your local bookstore, compare prices or contact your local distributor.

Sample Chapter

This is a free version of pages 45, 46, 47, 53 and 54 of Information Architecture For Designers.

Spanish translation of this sample chapter: Arquitectura de información para diseñadores: diseñar sitios web para el éxito del negocio - capítulo 2

This page doesn't contain the illustrations and cartoons on these pages in the book. See the table of contents for more details.

Chapter intro: Audience Research

A common mistake when designing a website is to assume that the people who will use it are just like you. Humans naturally assume that other people are like themselves. The goal of researching your audience is to let you step into the shoes of the real users of your website.

Even though this research chapter is one of the longest in the book, don't get the impression that every design decision you make should be based on research. Experience and common sense are even more powerful allies. Doing research gives you insight into how other people use the web, and helps hone your instincts for what will work and what won't. Research creates experience. Experience means you can make informed decisions.

Research will make your life easier, not harder. Remember all those times you have had "which-features-would-be-more-useful" or "how-should-we-implement-this" type discussions? Highly paid teams spend hours arguing about how things should be done, and the arguments are based on little more than personal preference. Having a few research techniques up your sleeve will allow you to find an answer to those questions based on real data, not personal preferences. Doing research is quicker and more cost-effective than going through endless discussions in which not the best argument but the strongest personality usually wins, resulting in the wrong design decision being made for the wrong reason.

First-time researcher

"On one project, I called local hotels and set up 15-minute interviews with the concierge at each. It didn't take up any more of the budget than sitting at my desk guessing either, and was much more valuable." Garrick Van Buren, Information Architect

Audience research just means asking yourself some questions about your audience -something all designers do - and trying to find real answers to those questions. There are plenty of questions that influence design: What is the first thing people want to do on the homepage? What are the three things we want to make really easy for people to do on the website? What percentage of our audience has Flash installed? Answering them from experience is good. Answering them with research is better.

It is easy to build some formal research into a project for the first time. When you ask yourself who your audience is and who you are designing this website for, have a look at the research
techniques in this chapter and pick one. You don't have to set up a huge research plan; just spend a few hours on it. You will always learn something interesting, and before you know it you will be doing research for every website you build.

Asking the right questions

"We're starting a project this fall for a FTSE 100 company that won't go ahead until research is completed outlining the direction that it should take this year. We suggested researching their audience last year and it worked so well that they want to do it again, ensuring that we're still moving in the right direction. Nothing replaces experience. However, some old-fashioned research helps to solidify your hypotheses almost every time. We've never found research to be a waste." Jared Folkmann, Information Architect

To do research, you first have to decide what questions you want to answer. Asking good questions is just as important as coming up with answers: answers close the mind, questions open it. When you think you know the answer, you stop looking. Try to answer the wrong question and you are wasting valuable research budget. Focus on formulating the right question first.

The user research questions relevant to web development tend to fall into one of these categories:

• Who is my audience?
• What are their goals?
• What are the best ways for them to achieve their goals?

Generic questions like these should be asked at the start of building of any website: what exactly will people try to do on the website? How will they do it? But websites are more often redesigned than built from scratch. In those cases, the questions to answer tend to be more practical.

Below are some typical practical research questions. You don't ask these of the website's users - you find ways to answer them by talking to the client and users of the website. Sometimes these questions are given to you by the client; sometimes you have to think them up yourself.

Question: "We are considering adding a certain piece of new functionality to our website, but is it worth the expense?"

First, find out what business goals this new functionality is trying to support - why does the client want it? Talk to the client: what positive effect do they expect this functionality to have? More people visiting the site? More sales? Try to put numbers on it. You can then calculate the expected profit (actual profit is usually lower) against the expected cost (actual cost is usually higher), and decide if it is worth building.

Next, talk to users and find out if the new functionality addresses a real user goal. Often, clients imagine users are having certain problems, when the real problems lie somewhere else. Building functionality to address the wrong problems is a waste of money. Use interviews or other research techniques described in this chapter to find out if this functionality addresses real user problems, and, if not, what the real user problems are.

"I almost always do some user research, and usually commit a good proportion of my resources to it. When I have had to do without, I worry that I have made the wrong assumptions about the users and may be designing something that they cannot use." Donna Maurer, Information Architect

Question: "How can we get more people to join our mailing list?"

Again, find out what the real question is first. Does the client want to get more people coming to the website through this mailing list? Or sell more advertising on the mailing list? That is your business goal.

Next, talk to users and observe them signing up. This way, you can identify obstacles to signing up, whether they are on the sign-up page or in the user's head. A usability test (explained later in this chapter) will show if the subscription page is hard to use - a surprisingly large number of subscription pages are. Interviews can uncover further confusion that users may experience, or problems with the perception of the mailing list.

Question: "We have lots of technical support documents - how can we best put them on the web?"

Let's assume the client has no existing website - so there is nothing up there to test. You need to find out how users will want to use these documents. Will they look for documents by product name, product identifier, or problem category? Chapter 3, page 76, discusses how to design an appropriate information architecture. You can start by examining the documents. Then interview customer support personnel. What are the top 10 needs of people contacting them?

When given a research question, you often first have to uncover what you really need to find out. Work on the question before you start working on the answer. You will usually end up with more than one research question.

The following text comes from the same chapter, from the section on contextual inquiry - p. 53 and 54.

Observation with contextual inquiry

Observation is the best way to find out how people actually do things. Contextual inquiry is a specific technique that combines observation with interviews. It leads to detailed insights about how people do complex things like buying a car, choosing a loan, or staying in touch with their fly-fishing buddies. Once you know exactly how people do those things, what their problems and specific needs are, you can support those tasks on your website.

The "contextual" part is crucial - it means that you have to go to where people are currently doing these tasks. Yes, get out of your office. Observe people where and when they do things for real: usually at home or at work.

Contextual inquiry tends to take time - the time for transportation alone will eat up lots of budget. A typical contextual inquiry exercise will take at least a week: a day for planning and finding people, two and a half days of observation and interviews, and a day and a half of analysis.

Let's look at an example: say we are building a website to help people manage their personal finances: bank accounts, bills, investments, and so on. Contextual inquiry can help us to answer questions like: What are the main problems people have when dealing with their personal finances? What (if any) are the stages people go through when dealing with personal finances? Do married people do this differently than single people? And so on.

To do contextual inquiry, go and observe people dealing with their finances in their natural environment: probably at home. Find some people in your target audience, and explain to them you are doing research for a website that wants to help people manage their finances. Ask them when they normally do their finances, and ask if you can come over and observe them doing this.

People often don't understand what you're trying to do when you ask to observe them. They might think they're being tested, or that you're going to report on how well they do to their boss. Sometimes they think you are there to explain to them how they should do things, and they feel angry about that.

It's your responsibility as a researcher (did I mention you're a researcher now?) to make them feel good about the process. One way to explain this is to explain the relationship as a master-apprentice relationship. They are the master and you are the apprentice: you want to learn how they do things by observing them in action.

Researching people is fun, but you should always be aware of possible ethical implications and think about the feelings of the people you are researching:

• They might feel obliged to answer you.
• They might feel intimidated by you.
• They might worry about what the things they said will be used for.

To allay these fears you should:

• Clearly explain to them what the research will be used for.
• Clearly explain to them that they can stop at any time during the research if they feel uncomfortable for any reason.
• Have them sign a consent form that outlines their rights before starting.

Also, make sure you emphasize that you're not testing them - you're there to learn how they do these tasks. Observe them doing what they do, and ask them questions about it when something is not clear: "Why are you doing that?" "Do you usually do it like this?" Take a notebook for the observation, a camera (a cheap one will do) and something to record interviews with.

Contextual inquiry as a technique is such fun that it is easy to forget to spend time doing analysis: i. e. going over your notes and thinking about what it all means. When doing analysis, try to identify goals, motivations, and tasks, and look for insights into how your website can better support these.

Contextual inquiry is a very powerful technique in really understanding the needs, goals, and tasks of your target audience. It can lead to insights into the hidden goals and needs of your audience tha