HTML MASTER CLASS /// LEARN TAGS /// BUILD STRUCTURE /// SEMANTIC WEB /// HTML MASTER CLASS /// LEARN TAGS ///

Discovery Research

Two research tools that answer different questions: why users behave the way they do, and how many of them do it.

⚡ Total XP: 0|💻 management XP: 0

Skill Matrix

UNLOCK NODES BY LEARNING NEW TAGS.

Select an unlocked node to view details root

🚀 LEVEL UP TO SENIOR:Unlock 500+ Advanced Practical Challenges & Exercises.

🎓 COURSERA PARTNER:Earn professional Google, Meta, and IBM certificates to supercharge your resume.

Talking to five users can tell you what's broken. Surveying five hundred can tell you if it's worth fixing. Confusing the two is how roadmaps get built on anecdotes.

1Interviews: Optimize for Stories, Not Opinions

The goal of a user interview is not to collect opinions ('Do you like this?') but stories ('Walk me through the last time you did this'). Opinions are cheap and biased by whatever mood the person is in; a specific, recent story is evidence. Keep interviews to 5-8 users per round — past that, you're mostly hearing repeats, a sign you've reached saturation.

2Surveys: Optimize for Sample Size and Neutrality

A survey is only as good as its distribution and its wording. A 15-response survey sent only to your most engaged power users isn't representative of anything. Aim for statistically meaningful samples, randomize where you can, and pilot the survey on 3-4 people first to catch confusing or leading questions before they poison your whole dataset.

3Step-by-Step Breakdown

Introduction. User interviews and surveys are not interchangeable. Interviews are a qualitative tool for depth — they tell you WHY a user struggles. Surveys are a quantitative tool for breadth — they tell you HOW MANY users share that struggle.

The Mom Test. Never ask hypothetical questions like 'Would you use a feature that does X?' — people are polite and will say yes. Instead, ask about specific past behavior: 'Tell me about the last time you tried to do X. What did you actually do?'

Designing a Survey That Doesn't Lie. Surveys fail when questions are leading, double-barreled ('Is the app fast and reliable?'), or open to interpretation. Use closed, single-variable questions for statistics, and reserve one open text box at the end for anything you didn't think to ask.

Knowledge Check. A user in an interview says 'Sure, I'd probably use a feature like that.' Why is this the least trustworthy statement in the whole interview?

→It's a prediction about hypothetical future behavior, which people overstate to be polite — past behavior is the reliable signal
→It means the feature should be built exactly as described, without further validation

Summary. Run interviews to discover problems and generate hypotheses. Run surveys to size those problems across your user base. Neither replaces the other, and both fail if you ask people to predict their own future behavior.

Pascual Vila

Frontend Instructor // Code Syllabus

Common Pitfalls & Errors

The Error //

Asking hypothetical, future-tense questions

// Wrong
"Would you pay $10/month for a feature like this?"

// Correct
"What's the last tool you paid for to solve this problem, and how much was it?"

The Solution //

'Would you use X?' invites polite optimism, not truth. Anchor every question to something the person has actually done, and let them describe the last real instance in detail.

The Error //

Sending a survey with double-barreled or leading questions

// Wrong
"Do you agree the new dashboard is faster and more intuitive?" (2 variables, leading)

// Correct
"How would you rate the loading speed of the new dashboard?" (1-5 scale)
"How would you rate how easy the new dashboard is to navigate?" (1-5 scale)

The Solution //

A question that bundles two variables ('Was the app fast and easy to use?') makes the answer impossible to interpret. Split it, and strip out language that signals the 'right' answer.

Lesson Glossary

[01]The Mom Test

A set of rules for asking questions about a person's life so even your mom would give you honest, useful answers — focused on past behavior, not opinions or hypotheticals.

Code Preview