5:10 Justin TV vs Twitch, build for yourself vs built for others
10:43 Minimum remarkable product
16:45 Why you need metrics as a baseline
17:56 Log the 5-7 most important metrics to your business
22:10 Why you should fix one issue at a time
24:47 Talk to users to HAVE product ideas,  not the other way around
25:40 Consensus building after talking to users
30:40 Emmett on talking to users
33:23 Looking at metrics, explaining why your metrics improved
37:30 Solving seemingly conflicting customer problems
39:00 Invert your assumptions, question them
42:00 Product market fit
46:20 How Emmett found Twitch's most important customers

"if it's not for you, your intuition is lying to you."

Main Takeaway
1. When building MVP, get it out AS FAST AS YOU CAN. Thats the only way that you know, whether you are right or wrong.
2. If you proud of your MVP, you are too late.
3. Collect your user activity.
4. If you don't have time to build the analytics tool, use 3rd party software and then just Log the top 5 important activity metrics
5. Always talk to your users first, then build the feature. if possible don't do it backwards.
6. Don't ask for the users on what feature to build. Just listen to their problem and try to solve it.

Great insights from the creators of my biggest time sinks.

This is great advice it helps me a lot.

Great!Learn a lot from this class!!

Extremely enlightening class!

Between all the various YC videos and Stanford talks, this one ranks in my top 5. Very well spoken guests! It also helps when I am intimately familiar with the products they are referring to.

Great talk. Thank you all for your contribution. Priceless. Witty jokes make good vibe)

Absolutely well done and definitely keep it up!!! 👍👍👍👍👍

I found this session to be very dense with knowledge. I took 3 pages of notes. I was averaging two pages of notes for other lectures.

11:50 this was me recently. My last company we were able to build a landing page that said we are going to build this give me your email address and that worked perfectly.

This company I’m working with now we tried that and we got zero email sign-ups. It’s because this type of online product needed an actual interface that somebody could play with. One or two or three features.

Interesting how one thing works so well And one scenario but doesn’t work at all in another. A constant learning exercise.

Great interview i learn a lot about how to grow up a producto and lear more from costumers or users

Has anyone experienced of talking to the users who did not choose you? In my experience, when they don't choose you, they don't care, and it's hard to talk to them. Any experience is welcomed here.

This is a great video!

awesome, must see! 👏

Thank You

Watching this after Emmett became OpenAI's new CEO

So good

what are the habits Steve refers too when he is talking about creating products customer like/dislike? 20:55