1/ The first 18 months of a startup:
After starting my 2nd company in 2019, I decided I would write down useful lessons I learned or re-learned along the way. Some were hard-earned & others required steady focus. A thread that I hope may help other founders starting out 👇
1/ I became "CEO" at 20. I dropped out of college. I had only interned somewhere prev. Looking back, I couldn't imagine the journey that would occur from writing code all day to scaling to 300 people. I got lucky, I screwed up a lot, & had a lot of help. Here's what I learned...
1/ Let's talk about fundraising: In 2009, Y Combinator only gave us $15K, half our YC batch died after demo day, 11 firms told us no, 2 firms wanted to find us a CEO, 1 VC told me, point blank, we would fail, we were a week away from death but luckily raised $500K @ 2M pre.
There's no doubt in my mind that making AI art is real art.
I spent about 1.5 hours tweaking things to produce these visuals. The hair, perfect red lipstick color, focus, eyes, wrinkles, theme, reflections, clothes. It was a joy to achieve a result I couldn't have previously.
Imagine VCs take 80% of your company for funding, make you payback the funding through your first profits, and all they mostly do is help you find some hosting for your website.
That's basically the music industry for artists. One day there will be an unimaginable reckoning.
Twitter Spaces > Clubhouse. It just works. 23K+ people in a space listening to the President of El Salvador.
Well done
@jack
- you beat the odds, well.
1/ A bit of news: last week I decided to stop working on Mighty after 3.5 years 😓. If anyone is interested in buying the IP, please reach out.
This week our team will begin work on making new kinds of creative tools using advances in AI. A new kind of Adobe Creative Suite.
You don't need to pick anyone's brain anymore about startups. This is 10x more information than I had when I started out:
The onus is on you! You got this. 💪
I think people do not fully grasp how competitive the global talent market will become with remote work soon.
It is going to become irrelevant to migrate to the US. SV hacked around H1Bs and arcane visa policies.
When I was 20, I coded through new years eve because I felt stuck in life. I launched Mixpanel on Jan 1. Nobody noticed but I had a deadline to hit. I salute you hackers tonight 🫡
I am fairly certain most people cannot have their best ideas if they are in back to back meetings every day. You need time to question things and then research it.
So, if you wonder why it’s hard to make ground breaking products after the first, look at your calendar.
I once emailed Jensen cold last year and he got a driver bug fixed in 12 hours. 30 years after its founding, this guy still acts like a hardcore founder.
Jensen is the best tech CEO in the game rn.
In year one you likely don’t need:
- A fancy office
- A full-time assistant
- An office / ops mgr
- 2000+ sqft of office space
- Fancy furniture
- Lots of employees (> 10)
- Perfect design
What you do need:
A product people love in spite of its flaws.
The rest will come.
1/ Starting my road to learning about AI. I’ve done some courses here and there but I am planning a more consistent effort over the next few years.
Mostly documenting my path for others and for fun. Often people suggest things I didn’t know which can be helpful.
I am still willing to bet that Ethereum will significantly eclipse Bitcoin long-term. It's one of very few things that makes the Internet feel exciting again.
Apparently Google has released a ton of new ML/AI datasets in the last year or so.
If I was going to start a new company, I’d start experimenting with these for inspiration. This is just so cool:
1/ Introducing AI-first image editing to Playground—a way to instruct an AI to synthesize spectacular yet subtle edits
Try it here:
We are blown away by the creative possibilities with this.
Example: "Make it summer"
(More examples in a thread)
Startups are the best way to meet all your heroes in an industry. People who get shit done are almost always willing to meet, dream, and give away secrets about the future—so long as you’ve built something cool too.
It’s quite frustrating to feel so behind in software right now after so intensely being a student of it since 8th grade. I am not sure how to catch up other than just continuing to put in the work to accumulate AI knowledge. So if you feel that way, you’re not alone.
1/ The first 18 months of starting a company is often life or death. I must've made 5 different companies that each failed within 9 mo. 😭 Each time the company failed I figured out what I could do better. Eventually startup
#6
got to $40K/mo by month 18. Here’s what I learned...
Founders: we need the best engineers to work on our hardest problems!
Engineers: I want to work on AI, interesting hardware problems, and scaling!
Reality: hey, so can you make a landing page, make forgot about password, and fix these 3 bugs customers are complaining about?
Startup vanity: Valuation, Press, Funding, # of Employees, Offices, Meetings w/ famous people, Meeting to give lots of advice
Success: Revenue, Users, Word of mouth, Retention, Servers on fire, Emails asking for features, Meetings w/ users
1/ Today we're unveiling Mighty: a faster browser that is entirely streamed from a powerful computer in the cloud.
Demo in the next tweet 👇
New website:
Hacker groups are now dropping ransomware but not requesting money. They're requesting a withdrawal of Russian troops from the Ukraine to add pressure.
Problems remote work has that need to be solved:
- Loneliness
- Pairing quickly on problems
- Whiteboard brainstorm sessions
- Team comrade
- Quick drive-by conversations (sometimes bad too)
There's so much opportunity for this next revolution. It's not just about presence.
FanDuel raised $416M at a $1.3B valuation & got acquired for $465M. The founders will get nothing after nearly 10 years--can't really beat pref stack. Maybe they took something off the table but still a farcry from the goal. Let this be a reminder: raise what your company needs.
Things I’ve used to start a new company:
- Discord instead of Slack
- Google domains
- Still using G-Suite
- Still using Github
- More Node & TypeScript vs Python
- Typeform is amazing!
- Heavy google docs users
- Github static pages for a quick landing page
- Avoid g-hangouts
'The hardest part in a startup is that you wake up one morning, and you feel great about the day, and you think, "We're kicking ass." And then you wake up the next morning, and you think "We're dead." And literally nothing's changed.' -
@jkraus
You can make someone a 20% better engineer but you can't make someone genuinely care about the product you're making. That's why I'd rather work with someone worse technically that wants the product they get to help build.
At long last, we finally reached a milestone we set out to hit 18 months ago: 10 happy users who don't want to go back to Chrome, use the product for hours every day, and pay for it.
I know it's a tiny number but we ruthlessly focused on making just a few people happy first.
There's no upside in hating on someone's idea. You come off as non-believer or ruin their confidence. Instead, just ask the questions that make you hesitate & let them come to their own conclusion. If you're right, you helped them think it through. If you're wrong, you'll learn.
The Google AI team is in such a tough place: the targets will keep moving as their competitors advance. Any launch will hurt the brand that doesn’t live up to expectations which are rising. Throwing thousands of engineers to accelerate typically will slow things down. It’d be
In summary, 600+ comments later:
Excercise
Eat healthier
Guard your time
Start a family sooner
Work on sleeping better
Care less about what people think
Don't buy stainless steel finishes in the kitchen
One of the less talked about realities of achieving something great is the sheer amount of time you have to be alone, toiling until the thing you’re creating is great. Much of one’s perseverance is believing you’re not wasting your time despite feeling lonely w/ many distractions
The first 10 paid users took 18 mo. Getting to 50 took 6 mo. I think we will be at 100 in less than 2 mo.
This all takes a lot of time, iteration, and grit.
I think the most clever thing to do right now is watch how teens use ChatGPT vs the fringe business use-cases tech people use it for. The emergent use cases from technology often come the people who (1) have a lot of time and (2) must outsmart authority for freedom.
If you are confused why things aren’t headed in the right direction for your startup, I’ve found it’s a trap to start with your metrics, spreadsheets, surveys, or tickets.
Start by talking to tons of users face to face. Then match the qualitative to the quantitative after.
I am pretty sure the biggest difference between a startup and a large company is the anxiousness one feels if they haven’t shipped something meaningful in a week. In a large company it is totally acceptable for an 6+ mo project to not ship. That cannot happen in a startup.
The road to product market fit can be summarized as:
- Ship new code
- Have users try the product
- Frantically fix bugs
- Ask for brutal feedback
- Monitor use till they churn
- Learn why & write it down
- Prioritize what prevents churn
- Write code
- Find new users
- Repeat
Startups vs 1K+ company:
Two co-founders: “Hey, what you do think of doing X? Sounds good but maybe tweak it by doing Y?”
Bigger co:
Write internal doc
Send doc to mailing list
Everyone from mktg to legal chimes in
VP reviews
Features get added
Re-org occurs
Ships 1 yr later
Listen to this AI generated song featuring Drake & The Weeknd.
It goes so damn hard.
It's by "Ghostwriter977" on TikTok and it's blowing up on socials + streaming platforms.
UMG, which controls around 1/3 of the global music market, has already asked streaming platforms to ban
1/ Getting my first 100 customers always felt like a puzzle. The next 1000 seemed unreachable. Besides, how can you get feedback to make the product better w/o users? After many years, we ended up w/ 6,000+ paying customers. It was a grind to get there.😩 Here's what I Iearned...
If I was a young or a college student in need of a little income per month, now is a brief window of opportunity where you can make really easy apps dedicated to solving one problem using ChatGPT in a single step where many back and forth chats are required. Make the UI visual.
My algorithm for deciding how much to raise was simple:
- $225000 fully loaded cost per hire (SF)
- times 6 people (max # I'll need for now)
- times 2 years of runway (time I need to succeed)
- times 1.3x for a buffer (because I'll get something wrong)
Raise just what you need.
The three dashboards you need at a startup:
- stability dashboard: does the product do what it says w/o breaking
- product dashboard: 2-3 things that prove people use the product & come back to use it again
- revenue dashboard: you make money to be sustainable & can grow
One reason you cannot easily part-time do a startup is that so many of the best, clever ideas occur somewhat spontaneously. You sort of look at all the possible angles over a span of 12-18 months and one day, while looking at a tree, the proverbial apple bonks you on head.
My favorite question to ask investors at the end of a pitch is: "What do you suspect the reason we will fail will be?"
Benefits:
(1) You get *some* feedback (don't be defensive)
(2) You can learn if they know something you don't
(3) Counter-interview - are they smart too?
This week I've witnessed some of the most rude, mean spirited, dismissive people. I suspect IRL they're much different.
Meanwhile, there's been an outpouring of support, encouragement, and optimism to counter-act it all.
My team & family appreciate it. All fuel for fire. 🙏🏾
A small team that knows how to work together should be fiercely protected. Even adding a single semi-misaligned person can drag the team into endless delay and steep deceleration.
I suppose I've learned today how many people are excited to see our technology industry fail. It's unfortunate that it went from something exciting and futuristic to something political where people root for it to fail--even if others are collateral. I guess we broke something.