Mara logo

User Research

The systematic study of users' goals, behaviour and context — through interviews, observation and analysis — to inform what gets designed and built.

Process & Methods

Research splits into generative work, which discovers what problems exist and is done before deciding what to build, and evaluative work such as usability testing, which checks whether a specific solution works. Both are behavioural: what people did and why, not what they say they would do. Stated preference is a poor predictor of action, which is why interviews probe past behaviour rather than future intent.

The economics are the argument. A week of interviews that stops a quarter of misdirected engineering is the cheapest thing on the roadmap. The common objection — that there is no time — usually means research is imagined as a three-month study, when five interviews in four days answers most questions well enough to act on.

In practice

An AI tool was about to add a chat interface because competitors had one. Five interviews revealed users did not want conversation, they wanted an audit trail of what the model changed. Before: a chat feature was scoped. After, a diff view shipped instead and became the most-used screen. The catch — the chat feature was already in the investor roadmap.

Learn more

You may ask

Frequently Asked Questions

How much research is enough?

Enough to stop hearing new things. In practice five to eight interviews per audience reaches diminishing returns for most product questions; if every conversation still surprises you, keep going.

Can analytics replace user research?

No — they answer different questions. Analytics tell you what happened and at what scale; research tells you why. A drop-off in the data becomes actionable only once you know what people were trying to do.

Related terms

All terms

Let's talk about your product.

Happy to look at what you're building and say where design would move the needle.

Contact Us