[Notes] Do things that don't scale - Paul Graham

Hi all! I thought you might like a quick distillation of PG's essay. If this sort of content resonates with the IH crowd, I can share more notes I've taken on his essays and misc Startup School lectures.

Notes:

• Startups take off because founders make them take off.
• You can build the best glider in the world, but without a thoughtfully coordinated launch, it's guaranteed to go nowhere. Kickstarting a startup requires a lot of thought, time, and energy.
• Nearly all startups require initial users to be manually (and aggressively) recruited.
• Many founders resist this exercise because they're either shy or lazy. Writing code is easy. Sales and marketing are hard.
• Every successful startup started with zero users and had to work hard to get the first 10, and then the next 90, and then the next 900. Don't forget the power of compound growth.
• Start measuring your startup by weekly growth rate and try to maintain or increase it every week.
• Marketplaces and platforms require heroic efforts to gain traction, so expect that you'll be high-touch for a long time. Get good at being high-touch. Learn to love it. Use tools that make it easy to love it.
• Startups are like human babies: they're super fragile and they need really good parents. Good startups have good parents that are willing to do tedious, laborious, painful things. Whatever it takes.
• The optimal path for dominating a market usually begins with manual recruiting.
• "How do you find users to recruit manually? If you build something to solve your own problems, then you only have to find your peers, which is usually straightforward."
• One way to figure out who to recruit is to get initial users with an untargeted launch and then target the most enthusiastic types of users.
• Constantly think about new ways to make your users smile. Write them handwritten notes, get to know them, cater your communication to their preferences, etc. Their delight should be your top priority - get creative!
• Early startups need their founders to offer insanely great customer service, attentiveness, and communication skills.
• Over-engaging with customers is not only good, it's required. It's what keeps the (critical-to-survival) feedback loop rolling.
• Start with a niche. If you're B2B, that means other startups. You're starting a fire - and before you can add more logs, you need to fuel it with kindling and keep it contained.
• Think of yourself as an unpaid consultant, and use your product to delight them. Even if it means using your product on their behalf, or even acquiring new users for them by using your product (if you're building for lead generation).
• You don't need to write any code in order to be a consultant. You can be the code: manually do the things your code would do. This lets you quickly validate or invalidate your idea.
• Launches don't matter at all.
• Partnerships usually don't work.
• The best strategy will fail if it doesn't have sufficient energy going into it. You need a heroic level of effort to make a startup work.
• The things that you do in these early stages - the things that don't scale - will become a part of your company's culture when it's big. If you obsessively delight users now, you'll probably keep it up later.
• You probably shouldn't hire salespeople before you've done lots of sales yourself.

How I'm using it:

• I'm working on PMAlerts on nights and weekends - I'm trying to figure out how to most efficiently convert my hours into MRR. The "unpaid consultant" bit stood out to me; I'm going to try to use PMAlerts on behalf of other founders to generate leads for folks before reaching out to them.