What is UX research?
UX research replaces assumptions about users with observations of them. It asks what people are trying to do, how they do it now, what gets in their way and how a design performs when real people try it. The answers come from two kinds of method. Qualitative research, such as interviews, usability tests, field observation and diary studies, explains why people behave as they do. Quantitative research, such as analytics, surveys, A/B tests and heatmaps, measures how many do what, and how often.
It also splits by timing. Generative research, at the start, finds the problems worth solving and the needs a product could meet. Evaluative research tests whether a particular design solves them, from paper prototypes to the live product. A good research program does both, and feeds its findings straight into the work rather than into a report on a shelf.
Why it matters
Why UX research matters
Every team designs for an imagined user, and the imagined user is usually the team. Research is how a brand discovers that customers call the product something different, look for the price in a different place or abandon the form at a field nobody thought twice about. Finding that before launch costs a few interviews; finding it after costs a redesign and the customers lost in between. Research also ends arguments between opinions, because a recording of five people failing the same task is more persuasive than any slide. It is also faster and cheaper than most teams assume: a week of interviews and recorded sessions is a small price next to building a feature nobody uses.
In practice
How UX research works in practice
- 01
It starts with questions, not methods. What do we need to know, and what decision depends on it? A question about why people abandon a booking calls for interviews and session recordings; a question about which headline works calls for an A/B test. Participants are recruited to match the real audience (existing customers, people who chose a competitor, people who have never heard of the brand), and sessions are planned around realistic tasks.
- 02
Small studies go a long way. Five or six usability sessions per round reveal most serious problems, so testing early and often beats one large study at the end. The team watches the sessions, ideally live; findings are written as short, prioritized problems with evidence, and each one is tied to a design change. Then the change is tested in turn, and analytics confirm whether it worked at scale.
Watch out
What people get wrong about UX research
Asking people what they want instead of watching what they do. Users are reliable about their problems and unreliable about solutions, or about predicting their own behavior. Second, testing with colleagues, friends or the client's own team, who already know too much. Third, doing research once, at the start, then building for a year without checking. And the expensive one: commissioning research and then designing what had already been decided. Findings also die in long decks nobody reads; a short list of problems the team can fix this week does far more.
