Image Image Image Image Image Image Image Image Image Image
Scroll to top


Improve Product Design: Watch What End Users Do

By Jonathan Lupo

The beginning of any design and development project should include activities aimed at a deep understanding of likely user behavior. Planning a design that facilitates likely user tasks will succeed in minimizing frustration and reducing user abandonment.

Recruiting the right types of representative end users is crucial to the process of planning a successful research initiative, known as collecting the voice of the customer (VOC). Another important facet of a successful research endeavor is a well-planned line of questioning.

Ask Open-ended Questions

When attempting to understand primary tasks and daily routines of likely end users, a line of questioning should be open-ended. Open-ended questions allow users to speak freely and naturally about their real experiences. When attempting to uncover the truth about a user’s online habits at the office, for instance, questions such as “Can you walk me through a typical day at the office?” are better than the question “Do you like to go online at work?” A robust response from a user allows for more observation than a simple “Yes” or “No.”

Ask Users to Recall, Not to Speculate

One rule of thumb when conducting an interview is to ask users about past experiences rather than asking them about their preferences. Many mistakes are made when the interviewer asks users to speculate or “brainstorm.” Questions that cause speculation often do not accurately reflect typical user behavior.

An example of a line of questioning that could cause a user to speculate would include a question such as, “What do you wish you could do on your online banking website?” If the goal is to accurately understand how to build a more useful website, it is more important to find out what the users need, not what the users say they want.

In order to uncover the truth about what the user needs, it is better for a facilitator to ask users to go to a website that they would typically visit, and ask users to recreate a recent visit. Users should then attempt to perform the tasks they tried to perform on that recent visit.

Do Not Introduce Bias

The point of the end-user interview is to access real user behavior and thinking, and not to influence or cause unnatural user responses. To avoid introducing facilitator bias, it is important to understand some simple facts about research participants before they enter into an interview session.

  1. A user may want to be helpful: Often, research participants will walk into a session with the innocent intention of wanting to be helpful. Asking the users questions that prompt them to speculate about web tactics, for instance, may cause this type of user to come up with what they believe to be great ideas for the facilitator. The reality may be, however, that the ideas brought up in this impromptu brainstorm actually may not be useful to the user.
  2. A user may wish to impress the facilitator or a peer group:One of the dangers of interview situations that involve multiple participants, such as a focus group, is that users often want to impress their peers. This psychological drive manifests in two ways. First, a group leader generally emerges within a focus group. That group leader may speak first, talk over others, influence shy individuals, and craft responses in a manner that makes the group leader seem more intelligent or important. The same phenomenon may happen in an interview with a single participant. In order to avoid verbal responses that seem inspired by ulterior motives, it is better to put end users in situations where they are performing tasks, recounting past events or doing both. It is easier to get an idea of likely behavior from observation of current behavior or an honest account of previous behavior.

Note Patterns of User Behavior in Personas

A designer can gain valuable insight by observing a user’s natural behavior. For example, in web design, the researcher may observe whether a group of end users access a website by searching Google or whether they attempt to perform a certain type of task. If multiple end-user representatives exhibit a behavior, a pattern may be established. Behavioral patterns and common tasks are generally aggregated in the form of a behavioral persona. A persona is an archetype, or a behavioral portrait of a typical end user. Personas aid in understanding the likely use cases of expected visitors to a website.

In the cases above, a design may need to support tactics aimed at optimization for external search, or a common task may need to be elevated in a design or through the information architecture of a website.

Watch and Learn

With any end-user interview, data gained from observation may be more valuable than answers provided by users. For web development, for example, an end-user interview should aim to uncover information about expected, online behavior. It is, therefore, more important to understand what users are likely to do on a website than what users tell the interview facilitator they are likely to do on a website.

About the Author:

Jonathan Lupo is vice president of information architecture at Empathy Lab, a digital strategy firm. He also is the author of a professional blog about information architecture, research and user-centered design. Contact Jonathan Lupo at jonathan (at)