UX Glossary · U
User research
Also: UX research
Systematically learning about the people who will use a product, so decisions rest on evidence rather than assumption.
User research covers the methods used to understand users, their tasks and their contexts: interviews, observation, usability testing, surveys, diary studies, analytics.
Two axes that decide the method
Attitudinal or behavioural. What people say, versus what they do. These diverge constantly and not through dishonesty: people are poor witnesses to their own habits.
Qualitative or quantitative. Why and how, versus how many and how much. Qualitative explains, quantitative measures prevalence.
Christian Rohrer’s landscape framing at Nielsen Norman Group organises the common methods on exactly these axes, and it is a useful way to check you are not answering a “how many” question with a method that cannot count.
Generative and evaluative
Generative research runs before there is a design and asks what the problem is. Evaluative research tests something that exists. Teams overwhelmingly do the second and skip the first, which is how you get a well-tested solution to a problem nobody had.
The most common failures
Asking people to predict their future behaviour, which produces confident and unreliable answers. Leading questions. Recruiting whoever is easy to reach rather than who actually uses the product. And running research after the decision has been made, which is not research but justification.
In practice
A small amount done early beats a large amount done late. Five conversations before a line is drawn will change more than fifty afterwards.