
Y Combinator
How To Talk To Users | Startup School
Summarised with Bite · 10 min read
Gustav from Y Combinator makes a simple claim that most founders still ignore: the best startup ideas do not come from isolated brainstorming, they come from repeated, honest conversations with real users. The talk matters because it shows not just why user interviews work, but exactly how to find people, what to ask, what mistakes ruin the signal, and how to turn messy feedback into a first product people might actually want.
0:01 – 5:44
Why real startups begin with real conversations
A photo says more than a pitch deck here. Gustav contrasts the Hollywood version of startup creation, the lone genius having a breakthrough, with a real Airbnb photo: Brian Chesky standing with Amal, Airbnb's very first guest, after they spent a weekend together in San Francisco. That image captures the core lesson of the talk. Great founders do not hide from customers until the product is polished. They learn from them before the product exists, and then keep learning for the entire life of the company. He sharpens the point with a story that sounds almost unreasonable. In 2010, Brian Chesky gave up his apartment and lived in 50 different Airbnbs over a couple of months. The stunt looked crazy from the outside, but Gustav explains the hidden logic: Brian got to talk to 50 hosts, day after day, and hear honest feedback directly from the people keeping the marketplace alive. Airbnb also put Brian and Joe's personal cell phone numbers on the website early on, instead of hiding behind anonymous support channels. That was radical then, and it still feels radical now in a world full of "do not reply" emails. The unexpected angle is that customer conversations are not a nice extra after growth starts, they are the thing that keeps a company honest, because users are the only stakeholders actually paying you. From there, Gustav moves into the practical problem founders immediately face: who should you talk to first? Friends and people in your network are easiest to reach, though they may soften the truth to avoid hurting your feelings. Former coworkers can be even better, especially if they know the domain well. And many founders eventually go beyond both groups, finding early users through LinkedIn, Reddit forums, Slack communities, Discord communities, and in person events. To make this concrete, he invents a mock startup around carbon emissions. His starting hypothesis is not a product idea but a problem space: companies want to reduce emissions, but often do not. So he targets founders, CEOs, CFOs, and people with titles containing words like carbon, climate, or sustainability, then reaches out with a short ask for a 20 minute call. The lesson is subtle but crucial: the first job is not to sell a solution, it is to locate the people who actually feel the pain.
2 more sections in the app
- 5:44 – 12:29How to run an interview without poisoning the answers
- 12:29 – 17:07Turning scattered notes into an MVP people might pay for
Summarised by Bite, an AI YouTube video summarizer that also works as a podcast summarizer.




