One of the most dangerous assumptions product developers can make is that their users are like them. This assumption is almost always wrong. While it may be obvious that a middle-aged product manager in a telecommunications company is not like the teenage users of the company’s mobile phone service; it may not be obvious that an electrical engineer developing a multimeter is not like the technicians who use the meter. Although the engineer and technicians have certain things in common, they differ in many critical ways. Their background and education is different. Their goals are different. The work context is different. And, inevitably, their knowledge of the product is different. The engineer has an intimate understanding of the multimeter’s specifications, its internal design, and the organizational environment that produced it. Technicians have knowledge on how to use the multimeter: in some cases this only covers how to complete the specific tasks required in their unique work context, in others it includes a thorough knowledge of the meter’s user interface, work-arounds for its limitations, and ingenious, novel applications for its use. When you consider that the multimeter has many types of users with varying backgrounds, goals, tasks, and work environments, you begin to understand the gap between product developers and product users. Closing this gap can make the difference between a modestly successful product and wildly successful one. User research helps to close the gap.
User research answers the following questions for actual or prospective users:
- Users and customers: Who are our users? If the user is not the product purchaser, who is? Who are our prospective users and purchaser?
- User characteristics: What are our users like? What is their age? Sex? Cultural background? Education? Training? Experience? Motivation? Physical limitations or disabilities?
- User goals, tasks, and models: What do users do with our product and what would they like to do? Are the tasks recreational or professional, easy or complex, variable or repetitive, frequent or infrequent, one at a time or multitasked, time critical or flexible, high-risk or low-risk, individual or collaborative? What mental model does the user have for their tasks and tools they use to accomplish them?
- Physical environment: What is the physical environment like? Noisy? Dusty? Hazardous? Dimly or brightly lit? Open or private office? Interruptions? Type of furnishings?
- Social environment: What is the social or organizational environment like? Social or professional peers? Work pressure? Organizational goals (explicit and implied)? Organizational structure and roles? Competitive or collaborative?
- User support: How available and effective is training, documentation, peer support, and technical support?
The questions above can be answered with a combination of direct field observation, indirect observation, interviews, and questionnaires. Customer support analysis and competitive analysis also provide useful data from which to develop UX requirements.
Direct field observation. One of the most reliable and effective ways to understand users is to directly observe them in the field, in other words in their work, home, or recreational environment. Field observation solves several problems inherent in other research methods: (1) Users’ work processes may be so automatic and behaviorally well integrated, that they cannot describe them. (2) Users environments may be so familiar that they are simply no longer consciously considered; users have learned to accommodate it so well they can no longer describe it. Even when users can describe their work processes and environments, without direct experience researchers may misunderstand their description. Direct observation avoids these pitfalls because it gives the researcher direct access to the pertinent information.
Field observation techniques are adapted from ethnographic methods used by anthropologists and sociologists that were developed to put those observed at ease. In fact, when implemented properly, field studies have significant PR value because they make users feel valued. Field observation is an excellent method to gather information on the users’ behavioral and sociocultural experience that will eventually help define product features and user interface.
Indirect observation is the collection of behavioral data without the researcher being present. Specialized software is available to facilitate indirect observation of user interaction with software, Web sites, and Web applications. Client-side applications can record and aggregate keystrokes, mouse clicks, and page views. Server-side applications record IP addresses of visitors and their mouse clicks (clickstream). This information can help researchers understand how users interact with their products; however, because it is difficult to interpret the intent behind the behavior, indirect observation usually must be combined with other methods to be useful.
Interviews help researchers understand the motivation behind user actions and are usually done as part of a field observation. Interviews can also be done independently, either face-to-face or on the telephone. Although some users may offer design suggestions and ideas during an interview, the purpose of the interview is to help the researcher understand what users need and how they work, not to get them to design the product. Typically, interviews are semi-structured, meaning some standard questions are asked, but new areas of inquiry that develop during the interview can also be explored. Special techniques can be used to get maximum value for the interaction; for example, users may be asked to sort names of products written on index cards into categories they feel are related, a method called card sorting. This information can be used to design an e-commerce site with products located where users expect them. Specially structured group interviews called focus groups can also be used to collect data and gage reactions to early prototypes. They are particularly useful to evaluate visceral responses and sociocultural influences.
Questionnaires or surveys cannot replace observation because they do not provide behavioral information, nor can they replace interviews because they are missing the rich verbal and non-verbal conversation from which much useful information is gleaned. Questionnaires can, however, collect large amounts of user background information and user-satisfaction data, and therefore are an effective complement to other research methods. Compared to mail or telephone surveys, online questionnaires are very cost effective and can be rapidly deployed. They can also be used to rapidly test visceral responses to Web sites, graphical user interfaces (GUI), and other products. Unfortunately, poorly designed questionnaires provide confusing or misleading data or suffer from poor response rates. Therefore, care must be taken to define questionnaire objectives, word questions effectively, and structure the questionnaire to maximize response rate and simplify statistical analysis.
Customer support analysis. Reviewing customer support records for UX issues and interviewing support staff is a cost-effective way to identify UX issues with existing products. For example, after usability analysts learned that Lands’ End Web customers were phoning the contact center to get advice on what size clothes to buy, they added online measuring charts and instructions, which reduced call volume significantly [74].
Competitive analysis involves an in-depth deconstruction of a competing product in order to find its UX weaknesses so that you can create an attractive and usable alternative. Competitive analysis may include direct field observation, indirect observation, interviews, and UX evaluation methods (see How is User Experience Engineered?on p. 13).
Data collected during the user research phase ensures that you are developing a product that meets the needs of real users. With this data in hand, the next step is to define UX requirements.
Work products: User, Task, and Environment Profiles. Notes taken during field observations and interviews together with indirect observation and questionnaire data are summarized in profiles of the various users, their tasks, and environments. These profiles can be used by product managers to help conceptualize new products or features, by product designers to make sure these features are successfully implemented, and by marketers to help target product marketing efforts. Profiles can take several forms and contain different information, depending on the particular project and design environment.
For example, a user profile may be presented as a user persona [26]. Here, user research data are used to construct several fictional users or “personas.” Each one is given a name, a photo, and enough descriptive detail to make them seem real. Products are then designed to meet the needs of the personas. The chief benefit of the persona approach is that it makes it easier for product developers to conceptualize their users, which is especially difficult in large projects when designers are not directly involved in user researcher.
Task profiles may take the form of a task analyses [121]; this is a detailed description of user goals, tasks, and actions. Goals are what a user wants to accomplish, for example, find a customer’s phone number. Tasks are how they plan to accomplish it, for example, search the database. Actions are the specific steps they need to follow, for example, open the search widow, type the customer’s name in the search box, and press enter. Detailed descriptions of this type can reveal opportunities to improve workflow and they can also accelerate user interface design by providing structure for design decisions.
User profiles may include a description of the user model [90; 91; 111]. A user model is a type of mental model—typically only partly conscious—that conceptualizes users’ work and technology. It influences user behavior and considering it during product design makes a product more intuitive to use, facilitating user acceptance.