Product5 min read

User Research

Conduct effective user research - user interviews, usability testing, Jobs-to-Be-Done framework, synthesis, and translating insights into product decisions.

User ResearchUX ResearchUser InterviewsUsability TestingJobs to Be Done

skill: User Research category: Product tags: [User Research, UX Research, User Interviews, Usability Testing, Jobs to Be Done] readingTime: 5

Most product failures come from building what users say they want rather than solving what they actually need. Good user research uncovers the underlying job-to-be-done, not the feature request. Users are experts at describing problems; you are the expert at designing solutions.

Workflow

Conduct user interviews correctlyUser interviews are the highest-signal research method for understanding problems, motivations, and current behavior. The key is asking about past behavior, not hypothetical futures.
Apply Jobs-to-Be-Done (JTBD) frameworkPeople don't buy products - they hire products to do a job. Understanding the "job" reveals the real competition and unmet needs that features can't capture.
Run usability testsUsability tests reveal where users get stuck - you can't predict usability problems from looking at your own design. 5 users find 85% of usability problems.
Synthesize research into insightsRaw interview transcripts are data, not insights. Synthesis identifies patterns across multiple users that reveal actionable product opportunities.
Translate insights into product decisionsResearch that doesn't change decisions is theater. Connect insights to product roadmap items with explicit reasoning.

Output Format

User research: past-behavior interview questions, JTBD statement template, scenario-based usability tasks, affinity mapping synthesis, and research-to-decision log.

Interview: "Tell me about the last time [behavior]" + "Why?" follow-ups
JTBD: "When [situation], I want [motivation], so I can [outcome]"
Usability: scenario task + think-aloud + observe hesitation
Synthesis: sticky notes → clusters → HMW statements
Decision log: decision + evidence + alternative + success metric

Examples

Input

Output

Edge Cases & Gotchas

Recruiting bias - Recruiting friends or power users creates selection bias - these are not representative of your average user. Use user testing platforms (UserTesting.com) or recruit from your broader user base by email.
Observer effect - Users behave differently when watched. Use think-aloud to externalize their natural behavior. Remote unmoderated tests sometimes reveal more natural behavior than live moderated sessions.
Research without action - If research findings consistently don't change product decisions, researchers lose credibility and budget. Close the loop: track whether insights led to changes, and measure outcome.
Generative vs. evaluative research - Generative research (discovery) asks "What are users' problems?" Evaluative research (usability testing) asks "Does our solution solve the problem?" Both are needed - don't confuse them.

When to Break These Rules

Skip formal user research for rapid prototyping sprints - build, ship to a small beta group, and measure behavior. Analytics is faster feedback than interviews when you have enough users.
Use surveys instead of interviews when you need quantitative validation of a hypothesis across a large sample - interviews are best for discovery, surveys for validation at scale.

Reference Files

https://www.nngroup.com/articles/which-ux-research-methods/
https://www.jtbd.info/