Paul Graham, Founder of Y Combinator, talks about how to get ideas for startups and six counterintuitive components of startups.

The way to get startup ideas is not to try to think of startup ideas. It’s to look for problems, preferably problems you have yourself.

The very best startup ideas tend to have three things in common: they’re something the founders themselves want, that they themselves can build, and that few others realize are worth doing. Microsoft, Apple, Yahoo, Google, and Facebook all began this way.

To jump to a specific point of the video, you can add the time parameter to the end of the youtube url. For example, 4 minutes and 15 seconds becomes: &t=4m15s
So to jump to point #1: https://www.youtube.com/watch?v=ii1jcLg-eIQ&t=4m15s

Notes

1:27 – Startups are counterintuitive like skiing
3:10 – Why do founders ignore partners’ advice
3:44 – lots of ski instructors not a lot of running instructors
4:15 – #1 – Intuitions about people you can trust
5:11 – work with people you respect
5:33 – #2 – what you need to succeed in startup is not expertise in startup, it’s expertise in your user
8:17 – “playing house”
10:18 – fundraising tricks, growth hacks (bullsh*t)
11:26 – #3 – Startups is where gaming the system stops working
12:57 – can fool investors but not in your best interests
13:25 – someone who builds something users love will do better than someone who knows fundraising
14:30 – #4 – startups are all consuming
16:10 – it never gets any easier
16:28 – student specific / university / do not start a startup in college / serendipity
21:28 – start at any age?
22:00 – #5 – you can’t tell if you’re up to the challenge of starting a startup
24:00 – no way to tell confident or quiet, same as military, only way to find out is to try (not in college)
24:30 – what do you need 1) idea, 2) cofounders
24:42 – #6 the way to get startup ideas is not to think of startup ideas
25:50 – turn mind into mind that sees startups: 1) learn a lot about things that matter (airBNB), 2) work on problems that interest you, 3) with people you like and respect
27:00 – how do you know if you’re working on real stuff – real problems are interesting
28:30 – gratifying your interest genuinely is the best way to prepare for a startup (and best way to live)
28:45 – edge of expanding technology represents opportunities (probably not steam engines, maybe)
29:00 – “live in the future”
29:25 – Harvard grad student wrote own VoIP w/o startup
30:20 – learn powerful things in college, domain expertise
31:00 – starting a startup cure for curiosity – just learn
31:30 – q&a
31:30 – what does non-technical founder do – domain or sales
32:37 – entrepreneur in business school – no
34:00 – how manage first few employees
36:20 – are we in bubble?
38:18 – labs to spin off startups
39:10 – female cofounder advice when seeking funding
40:37 – what would you learn in college
41:35 – recurring systems that make you efficient (have kids, amount of done you do per time is high)
43:21 – when is good time to turn side project into startup?
44:03 – what to do when side project is sorta growing but not hugely – reading do things that don’t scale
44:40 – what kind of start ups not go through YC?
45:45 – how do you figure out what matters? – anything on edge of tech but don’t really know
47:07 – snapchat
47:20 – hiring people you like might lead to monoculture and blindspots

Additional Reading

• How to Get Startup Ideas by Paul Graham
• Excerpt from Steve Jobs’s 1995 interview with Computerworld’s Oral History Project – watch from 1:10:53 onwards
• Transcript

Series

• Next: Building Product, Talking to Users, and Growing a Startup
• Previous: Startup Teams and Execution
• About Series: How to Start a Startup

Sam Altman, President of Y Combinator, talks about the last two topics of startups: Cofounders and Execution.

Notes

Q&A

• trends – trust instincts of young people
• burnout – address stressors

Cofounders

• choose someone you know and trust
• work at cofounder rich company (e.g. facebook rich like Stanford)
• top companies had 2 founders
• yc funds 10% of single founder cos
• 2 or 3 founders best, few more ok.
• “relentlessly resourceful”
• tough and calm (as well as smart and technical- coders start software cos, media ppl start media cos)
• James Bond more than expert in particular domain
• know founder and early hires for years
• try not to hire
• airBNB spent 5 months looking for first hire, then only 2 for first year; bad early hire can ruin co
• time spent hiring: 0% or 25% ( not at all if not absolutely needed or substantial time, to get it right)
• source of hires: usually personal referrals for at least the 1st 100 or more works out best
• most of time aptitude more important than experience unless role requires it
• are they smart, do they GTD, do I want to spend time w/ them
• good communication skills; manically determined; pass animal test; would feel comfortable reporting to them
• equity: roughly 10% to first 10 employees (over 4 years)
• fire fast

Execution

• culture of co set by watching founders execute day-to-day
• ceo jobs: set vision; raise money; evangelize; hire and manage; make sure entire co executes
• founder: focus (you and co): say no a lot; set overarching goals. repeat them; communicate
• remote co-founding teams really hard
• intensity (from founder) (startups not good for work life balance)
• always keep momentum

Additional Reading

• Stupid Apps and Changing the World by Sam Altman
• Do Things that Don’t Scale by Paul Graham
• Transcript

Series

• Next: Counterintuitive Parts of Startups and How to Get Ideas
• Previous: Startup Ideas and Products and Why to Start a Startup
• About Series: How to Start a Startup

Sam Altman, President of Y Combinator, talks about the first two topics of startups: Idea and Product.

Dustin Moskovitz, Cofounder of Facebook, discusses why and why not to start a startup.

Notes

Idea

• limit to how much you can figure out before getting product to users
• great execution of terrible idea is no good
• bad idea is still a bad idea
• ideas do matter
• idea definition: size and growth of market, growth strategy for co, defensibility
• if works out: 10 year commitment
• plans are worthless but the exercise of planning, is valuable
• don’t need to have everything figured out but nice kernel to start with
• someday need to build business that’s difficult to replicate
• idea first, startup second
• best companies are mission oriented: more buy-in: founders, employees, outsiders
• derivative companies don’t excite people
• best ideas often look terrible at first (“portal-less search engine”, “myspace for colleges”, “sleeping on couches”)
• idea that turns into a monopoly (start with small market, take over, and then expand)
• market that is going to be big in 10 years
• small rapidly growing market better than big, slow growing market
• can’t create a market that doesn’t want to exist
• Sequoia Capital’s question: Why now? why couldn’t it have been done 2 years ago and why will 2 years from now be too late?
• best if building something you yourself need, or else get close to your customer
• idea easy to explain otherwise it’s too complicated
• best ideas very different from existing companies in 1 important way or totally new
• clones usually fail (x for y, social network for pet owners, etc.)
• college good environment for meeting cofounders

Product

• product = anything that interacts with customer (including customer service, etc.)
• great idea leads to great product leads to great company
• top tasks: sitting in front of PC, talking to users. (not: raising money, business dev, getting press)
• build something users love; talk to users
• easier to expand from product a FEW people love to A LOT of people love than from product that a lot of people LIKE to a lot of people LOVE.
• growth by word-of-mouth when people really love your product
• over the long run great products win
• start with something simple (even if your eventual plans are complex) because it’s hard to build a great product
• do one thing extremely well
• founder being fanatical; feel physical pain over problems with products
• recruit good users by hand in the early days (not Google AdWords)
• do things that don’t scale: http://paulgraham.com/ds.html
• build feedback engine (user feedback > product > show it > repeat)
• gotta do sales and support yourself in early days
• metrics: measure right things and those will grow (see Lean Startup book)

Why start a startup?

• much stress
• 10 year commitment if a success
• 5 year commitment if a failure
• can’t quit as founder
• Boss – everyone else is boss
• Flexibility – always on call, you’re role model, always working
• not same as small business entrepreneur
• myth: you’ll make more
• employee: dropbox ($10B -> $10M) facebook ($200B -> $200M)
• founder: uber pet sitting ($100M -> $10M); über space travel ($2B -> $200M)
• myth: you’ll have more impact
• can have just as big or bigger at late stage company because of team, distribution, tech
• why do it: can’t not do it; right person to do it; got to make it happen
• you need to do it; you’ll need passion
• the world needs it; the world needs you

Additional Reading

• Advice for Ambitious 19 year olds by Sam Altman
• Good and Bad Reasons to Become an Entrepreneur by Dustin Moskovitz
• Transcript

Series

• Next: Startup Teams and Execution
• About Series: How to Start a Startup