15 lessons from a $15B ceo
lunch with jason, ceo of discord, and what made it a great experience
lunch and learns can either be more stale than leftover’s meal when stuck in a slide show or lively and interactive scaling to a room of 20+ strangers.
this one was the latter. matt mochary hosted a week full of lunches and dinners with different founders he’s worked with, inviting founders he’s currently working with.
the format and experience was ⭐⭐⭐⭐⭐.
what made the lunch and learn so good:
- time to mix and mingle to start where everyone could meet everyone
- community of liked-minded people, where everyone was 1 degree connection from his team and many were founders in his portfolio
- he asked questions to set the stage for everyone so no prior knowledge was needed, but also made sure it wasn’t stuck in product pitch mode or rosey glasses of success
- a majority of the time was q+a. and the timing was such, where people could ask a follow up question. it felt like everyone was part of the conversation, even though all attention was directed at jason
- jason is an awesome human! sharp yet humble. good storyteller and also honest.
here are 15 takeaways from the lunch:
1. the initial chat tool spun out of a passion project.
he started a gaming studio and ended up spinning out the internal chat tool during gaming to see if others wanted to use it. they had two teams face off. a red v. blue team dynamic to see who could do what in a race against the clock. this had me thinking if a way to build tools was start with passion projects first, then start a/b testing tooling that came up organically from teams solving their own problems.
2. early stage ceos make decisions without full picture.
early on, as ceo, he realized there was never a way to get the full picture before making a decision. there were no clear results, no robust data, no resources for research, and no ability to arrive at consensus that could help him with early decisions. he had to define success.
3. show users the product, and ask what could be better.
they worked with a small group of users, around 20 people, before launching. they got feedback and learned what they didn’t know they didn’t know about the product before launching. for example, they didn’t realize that latency in voice channels was the #1 thing and had to focus on that.
4. figure out your own path to 100 users, then to 1,000.
the ‘real’ launch was from a reddit post. they got their first couple dozen users from a public post. they welcomed people through voice channels and answered questions directly. reminds me of this lennys newsletter.
5. use analogies for immediate understanding of the product.
he compared a majority of their users using discord like a virtual living room. friends hanging out url between being irl. students jamming between classes. so clear, and it stays true to their video gaming origins.
6. monetization comes after usage.
the other side of the virtual living room, is the large communities as a service. midjourney is one of the latest, which has a whole community using their gen ai image tooling. that founder tried other interfaces before landing on discord. he figured out how to support these larger customers later.
7. small but mighty teams to start.
he wrote a manifesto (shortened version here) to keep the team lean but effective. this was counter to the belief at other vc backed companies to scale and grow at all cost. this changed during covid when they had a surge of users and needed to build more support.
8. observe + align teams to fix scaling problems.
the blitzscaling for the team was hard to support. he realized he needed to call things that didn’t make sense initially to activate #8. he sat in on meetings and just observed. he would bring up the list afterwards to the team lead to align on.
9. habits make the person, not the ceo.
someone asked what his top habits were, especially around learning, and i liked his answer: “i don’t think these habits make a ceo, these are things i always did that happen to now help me as ceo.” he loves to learn, and he loves his work.
10. for branding, pick a name that people can easily recognize.
also, they could trademark ‘discord.’
11. when the platform scaled to 50,000+, they had to think like a city.
at the scale of their user base, they had to figure out city level dynamics to keep out bad actors like building a police force. they built a trust + safety team of 200 people to figure it out. they also built tools to empower server managers.
12. growth is not linear, but variable and a roller coaster.
up and to the right is the macro view over quarters and years, not days and weeks.
13. he loves learning from others so working with a ceo coach was natural.
he ended up working with matt mochary because matt had other ceos he had worked with with relevant background and experience to problems he was facing.
14. discord complements, not replaces in person connection
when asking him about hybrid communities (strong irl event base building first discord server), he mentioned that discord is really a glue to hold moments in between the times of meeting. friends use it when they aren’t together. rather than replacing irl, it complements it.
15. always be learning.
jason was full of curiosity. sharing stories from what he learned from others in his team, around him, working with a coach, and reading. this also made his answers approachable yet direct. he connected immediately with the room of early stage founders.
thanks Matt + the team for hosting this! and thank you Quinntin for the invite!
hey, i’m michael and i host different community events in sf including hardware meetups, hackathons, and art exhibits. i like architecting great experiences as well as enjoying ones that others host.
connect with me on linkedin to see what’s up next.
