YC Root AccessLecture 16 - How to Run a User Interview (Emmett Shear)
Emmett Shear on emmett Shear explains user interviews to build products people want.
In this episode of YC Root Access, featuring Emmett Shear, Lecture 16 - How to Run a User Interview (Emmett Shear) explores emmett Shear explains user interviews to build products people want Emmett Shear argues that many startup failures come from building without understanding real user behavior and shows how targeted interviews corrected that mistake at Twitch.
At a glance
WHAT IT’S REALLY ABOUT
Emmett Shear explains user interviews to build products people want
- Emmett Shear argues that many startup failures come from building without understanding real user behavior and shows how targeted interviews corrected that mistake at Twitch.
- He emphasizes that selecting the right people to interview (including competitors’ users and non-users) is as important as the questions you ask because different groups reveal different blockers.
- The recommended interview style focuses on current habits and motivations—avoiding feature brainstorming and product demos—to prevent biased, “horseless carriage” answers.
- Twitch’s interviews surfaced underlying needs (money, stability, global reach, ease of broadcasting) that were more strategic than the literal feature requests streamers voiced.
- He outlines lightweight validation tactics (cheating with extensions, collecting payment commitments) and internal buy-in techniques (recording interviews and replaying clips).
IDEAS WORTH REMEMBERING
7 ideasWho you interview determines the strategy you’ll discover.
Twitch prioritized broadcasters (content suppliers) because viewers followed content; interviewing only viewers would have produced a different, less leverageable roadmap.
Start interviews by mapping real behavior, not imagined features.
Shear’s demo interview stays on how notes are taken today (tools, switching between pen/paper, collaboration, review habits) to reveal true workflows and frictions.
Avoid feature questions to prevent biased, low-signal answers.
Asking “Would you use/pay for X?” invites politeness and “faster horse” responses; instead, probe motivations, constraints, and what currently blocks progress.
Compare users, competitor users, and non-users to find the biggest blockers.
Current users complain about annoyances they tolerate; competitor users and non-users reveal the issues so severe they prevent adoption and market expansion.
Translate requests into underlying goals, then build the real fix.
Streamers asked for specific chat/admin features, but Twitch focused on deeper needs like monetization, video stability, and global accessibility—often things no one explicitly requested.
Validate with the cheapest real-world test you can run.
Instead of rebuilding a full product, “cheat” (e.g., browser extension) to ship one incremental improvement; when possible, require payment to confirm genuine demand.
Record interviews to align the team and speed buy-in.
Playing clips is more persuasive than secondhand summaries, and recording also keeps founders engaged rather than burying the conversation in note-taking.
WORDS WORTH SAVING
5 quotesWho you talk to is as important as what questions you ask.
— Emmett Shear
Don't show them your product.
— Emmett Shear
The most interesting things you learn in interviews come from the, 'Interesting, tell me more.'
— Emmett Shear
Sales is this cure-all for this problem. Get people to put, give you their credit card.
— Emmett Shear
Find a way to cheat is what it comes down to.
— Emmett Shear
QUESTIONS ANSWERED IN THIS EPISODE
5 questionsIn the note-taking app example, how would you decide whether the “most important user” is students, IT administrators, or parents—and what signals would change your mind?
Emmett Shear argues that many startup failures come from building without understanding real user behavior and shows how targeted interviews corrected that mistake at Twitch.
You say to avoid feature talk early; what are your go-to prompts for uncovering “big blockers” without drifting into solutions?
He emphasizes that selecting the right people to interview (including competitors’ users and non-users) is as important as the questions you ask because different groups reveal different blockers.
Twitch heard many detailed feature requests from existing users but deprioritized them—how do you decide when an annoyance is actually worth fixing versus noise?
The recommended interview style focuses on current habits and motivations—avoiding feature brainstorming and product demos—to prevent biased, “horseless carriage” answers.
When interviewing non-users, what’s your preferred way to separate “I’m not interested” (no market) from “I’m blocked by X” (fixable adoption barrier)?
Twitch’s interviews surfaced underlying needs (money, stability, global reach, ease of broadcasting) that were more strategic than the literal feature requests streamers voiced.
What’s a concrete example of turning an interview insight (e.g., “video stability in Europe”) into a measurable engineering roadmap and success metric?
He outlines lightweight validation tactics (cheating with extensions, collecting payment commitments) and internal buy-in techniques (recording interviews and replaying clips).
EVERY SPOKEN WORD
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome