The crucial people to get your product started for the first 6 months are not who will be using it 3 years later.

If you just talk to who's easy to talk to, you're not really getting the best data.

Users think they know what they want, but you get the horseless carriage effect where you're getting asked for a faster horse.

Twitch launched in June of 2011, and our growth ever since has exceeded even my expectations, which were not small. A year and a half later, the community of broadcasters and viewers has multiplied hundreds of percent.

The crucial people to get your product started for the first 6 months are not who will be using it 3 years later.

That's actually one of the most disappointing things about doing user interviews and user feedback, which is why I think... people don't do it. You're going to get negative news about your favorite pet feature most of the time.

It's one of the most validating things you can do for a product is go out there and get them to commit to pay you up front.