What is the primary purpose of UX research before starting development?

WHAT THIS TESTS: Whether you treat user research as an engineering risk-mitigation tool or as a design-team luxury. Senior engineers are expected to know that the most expensive mistakes are building the wrong thing or solving an imaginary problem. The interviewer wants to hear that early research keeps product-development efforts aligned with true user needs rather than assumptions, and that you understand the difference between quantitative data showing what happened and qualitative insight showing why.

A GOOD ANSWER COVERS: First, de-risking development by validating real user needs before writing code. The NNGroup emphasizes that earlier research has more impact on the final product because findings can steer the entire effort. Second, illuminating unknown unknowns during the discovery stage. Field studies, user interviews, and requirements gathering reveal constraints and needs that affect technical architecture and scope. Third, distinguishing what from why. Analytics and logs show what users did, but research reveals why they did it. Data alone is not a great substitute for talking to people. Fourth, cost efficiency. Course correction is cheapest before development begins. Research during the explore or discover phase prevents expensive rewrites later.

COMMON WRONG ANSWERS: Treating UX research as purely aesthetic or saying it is only for picking colors and fonts. Calling it a nice-to-have that can be cut when schedules slip. Believing that A/B testing, analytics, or search logs fully replace qualitative user research. Another red flag is saying research should only happen after a prototype is built to validate a design. While testing prototypes is valuable, the primary purpose of pre-development research is discovery and alignment, not post-hoc validation.

LIKELY FOLLOW-UPS: How would you advocate for research time if your product manager wants to ship immediately? What research methods would you pair with analytics to understand a drop-off in a funnel? How do you balance research with technical spikes when scoping a quarter? When is it appropriate to skip formal research and rely on domain expertise?

ONE CONCRETE EXAMPLE: Suppose your team wants to add an advanced search filter panel to a dashboard. Analytics show users visit the search page but do not use existing filters. Quantitative data tells you what is happening, but not why. Pre-development user interviews and task analysis might reveal that users do not understand the filter terminology, that the panel is hidden behind an unclear icon, or that users actually need a completely different sorting mechanism. Discovering this before building the new panel saves weeks of engineering effort and prevents shipping a feature that looks powerful but remains unused.

Source: nngroup.com