Status: list of pointers and misc musings; still baking
General great sources of advice
- Paul Graham’s essays
- Stripe Atlas Guides -
stripe Atlas guides | Stripe Guides
- Elad Gil’s “High Growth Handbook”
- Ben Kuhn’s essays
- Esp good for technical IC ⇒ founder/manager
- The Great CEO Within
- How to win friends & influence people
- YC’s “How to start a startup” sequence (?) link
- (probably okay but I haven’t watched the full thing)
Meta thoughts on advice:
- Startups are more “riding a bike” tacit knowledge, less “book” knowledge.
- Relatedly: knowledge transfer happens by spending time with great people. Ideally, work directly with them. Otherwise: talk to them, read their writing, or study their trajectory.
- Conversely: These sources above are worth reading in full instead of skimming or summarizing with ChatGPT. Books & writing are more than just their high-level takeaways; they’re ways to self-program, self-hypnotize. Still lots of alpha in books.
- Also useful to re-read!
- Advice is very situation-dependent. Consider reversing all advice you hear. If we say 10 things, maybe 7 will be somewhat helpful, 2 will be net negative, 1 might kill your startup, you need to figure out which is which.
- You should interpret a lot of things I say as “this was helpful for Austin”, but you might not be very much like Austin
What to make
- Lean into what you’re really good at, what your unfair advantage might be, what feels mysteriously hard for other people
- Easiest to build something for yourself, or people like you
- I get a lot of motivation from working on a thing that makes people like me happy
- “Ideas are easy, execution is everything”
How to Get Startup Ideas
Talking to users
- Coming up with ideas: Don’t pitch people your idea, instead learn who they are
- “The Mom Test”
- “Talking to Humans”
- Improving your product
- “Don’t Make Me Think” on user interviews
- watch people and see what they do
- Marketplaces
- “Cold Start Problem” (?)
Building your product
- Move fast!!!!!!!
- Fast = iterate, learn, improve quickly
- Pick a ship date, commit to it, cut scope as necessary, focus
- Fast = impressive to your partners (customers, investors, hires)
- Field-specific, but: get good taste, practice a lot, always be shipping (?)
- Eg software: tech blogs in your field
- Eg events: taco tuesday
- Research your competitors, consider the tradeoffs they made, copy what’s good
Visual Design
- Sub-skill of building, can have high variation
- As with driving, everyone thinks they have good design taste but… most don’t
- Resource:
Butterick’s Practical Typography
- Resource:
adamwathan Refactoring UI
Finances & pricing
- Stripe Atlas Guides
- Patio11 on pricing — “charge more”
- Worry about the big picture
- Usually, raising revenue >> lowering costs
- Much more upside potential in revenue, you can only get costs down by a factor of 1x, but you could 10x or more on revenue
- Easier to cut costs since you control your spending but not other’s spending
- Fundraising ≠revenue: funds raised are a promise, revenue is value delivered
- Esp important for charities to get this straight
- aside: your project don’t necessarily want to over-optimize for money early on. should be willing to spend early on things that will compound, grow
- learning & skills, connections, impressiveness, network. cf “What are you getting paid in”?
- eg first Manifest lost $30k, second Manifest made $60k, third on track for $200k
Writing
Extremely leveraged meta-skill
- Copywriting for websites, forms: cut words
- Marketing, blogs, newsletter: List of internet writing advice
- Cold emails/text messages
- Internal docs, written culture
- Send “investor” updates!
Marketing
Marketing: broadcast what you provide & why it matters
- Your “email” list, or generally where you broadcast things
- Owning your channel can be super important, vs chasing Twitter/YT/HN virality
- Your website, landing page
- Focus on value delivered to them
- “Socials” — X, Reddit/HN, Discords/GCs
Sales
- The art of the cold reach out, Manifest as a case study
- Model your recipient, match their language, cite people & examples that matter to them
- 1:1 reach outs with people
- Take notes on the call imo
- B2C sales: Manifest conference reach outs
- B2B sales: Manifest sponsorships
- Have a CRM!
- Basic Airtable/Notion with name, contact, stage in pipeline
Intros
- [todo]
Working with people
eg cofounders, employees, contractors
Building up your team:
- morale
- mission alignment
- incentive structures
Resource: “Peopleware”
Basic structure of running a team:
- daily-ish standup
- 1:1s w/ cofounders and employees
- all-hands
Do retros after big things:
- What went well
- What could have gone better
- Improvements for next time
Community
Community is where people talk to each other, not to you.
- Can be a moat, flywheel. Manifold, LessWrong were good at this.
- Have a place for your fans to talk (Discord, your product itself)
- Elevate great people (moderators, volunteers ⇒ employees)
Events: Hosting events is a subskill, can sometimes be very useful. There’s
- “The chairs are where the people go” — super underrated, short
- “The art of gathering” — generally well regarded
Mentorship, coaching
- [todo]
Distractions
- Legal stuff
- Entity structuring
- Worrying about downside risk
- Parties, events, conferences
- Maybe if you’re getting customers there
- Virality for virality’s sake
- Can be useful proof point; but key question is “how does it convert”
- Advice (oops)