a while back, one of my residents asked me how they can get into startups, and i made him a list of general points to consider. throwing this here in case i need to send it to anyone else in the future. note that it's fairly basic, for two reasons. 1. i'm not a yc partner, everything i've learned has been fairly basic. 2. startup advice is generally the same throughout. i have yet to really hear anything concerning how to build a startup that has been incredibly surprising.

start with a problem. [1]

dig wells, not troughs.

pick a good direction. ideas fail, but if you're moving in the right direction, pivoting shouldn't be too difficult, or too costly.

again, pick a good direction because you will have to change your initial idea.

a startup isn't a startup because it's small or new or a failed business. a startup is a startup because it moves quickly. that's the only difference between us and them.

you're never pot committed. the great part about being a founder is that you get to pick what you want to work on. you get to pick the problem you solve, the approach you take, the way you respond to input. don't waste that edge by thinking you're pot committed.

you're often competing for things other than $$$. competing for attention, space within an outfit, convenience is likely much more difficult than competing for $$$.

barriers to entry. by building something hard, you buy the right to fail. similarly, if you build something easy, you have no right to fail.

think for more than a couple of days. i don't think its recommended to force ideas to work, but if you are going to force an idea, at least give it more than a couple of days. this is something you will spend a lot of time, energy, money, and health on. make it worthwhile.

you're probably not zuck, so focus on building something people need instead of building something people find cool.

pick good cofounders. go it alone, or go it with really, truly great co-founders. guys that are punters, that guys who don't let anything get between them and work.

startups are inherently chaotic and they're inherently volatile. big roadblocks will only kill you if you let them.

you'll never feel ready to launch if you think for more than five minutes. but ship it anyways, unless what your shipping is so bad that it actively drives users away. you only get one, two shots at best before users won't try you again.

if you're an undergrad, or even a masters student, you're probably not a domain expert in anything super technical. some are, but if you are, you're probably not reading this anyways. keep this in mind.

"i recently told applicants to Y Combinator that the best advice i could give for getting in, per word, was 'explain what you've learned from users.' " -pg

a lot of the best never raised a round.

[1] funny to list this first, still debating whether this is truly necessary. bbviously products need to solve a problem, but do the founders need to know exactly what problem their products solve? could you substitute insane traction for clarity of a problem, or could you accept that insane traction always means that products are solving a problem.

4/9/2024