Make It Happen
There’s a common failure mode where founders wait for the company to take off, as if it were a weather event that shows up on its own. PG puts it plainly in Do Things That Don’t Scale: “startups take off because the founders make them take off.” Nothing about a startup is automatic at the start. The momentum you’ll later mistake for inevitability is something you build by hand, long before there’s any machine to do it for you.
The trap is understandable. We learn what companies look like by studying the ones that already won, and by then they do look like they run themselves, all self-sustaining and scalable and humming along without much visible effort. So founders copy the end state instead of the beginning. The trouble is that the finished product makes a terrible instruction manual. What gets a company to takeoff rarely resembles what it does at scale, and it often looks a little embarrassing.
Recruit users by hand
The most common unscalable thing founders skip is recruiting users manually, one at a time. PG’s favorite example is Airbnb. The founders didn’t wait for hosts to sign up. They flew to New York, went door to door, signed people up in person, and photographed the apartments themselves because the listings looked better that way. Stripe did a version of the same thing that became known as the Collison installation: instead of telling a prospect to go try it, the Collison brothers would take their laptop and set it up right there while they had them. And Wufoo, once someone signed up, sent them a handwritten thank-you note.
None of that feels like leverage. If anything it feels like the opposite, a founder doing work that clearly won’t scale to a million users. But it’s really an intelligence operation dressed up as grunt work. Every user you win by hand shows you where the real complexity hides: which objections matter, which edge cases keep coming back, where the product bends in ways nobody would have mentioned in a survey. You earn the right to automate something only after you’ve done it by hand enough times to understand its true shape. Automate before that and you don’t build leverage at all. You just automate your own ignorance.
Go narrow first
The related mistake is going wide too early. Chasing the whole market at once is tempting because it feels ambitious. But focus is about the only real advantage a startup has over an incumbent, and spreading it thin across a broad front is the fastest way to give it back.
PG’s rule of thumb is to aim for a market so small that it sounds unambitious, then own it completely. Facebook started at one school. eBay started with, of all things, collectors of Pez dispensers and other odds and ends. The point isn’t the smallness, it’s that a market you dominate gives you a tight feedback loop and real fans, the kind of users who tell other people because you fixed something no one else bothered with. A small market you own outright beats a big one where you’re just another option. Expanding later from a position of strength is far easier than expanding from a beachhead you never fully took.
Build what you’d want to exist
The most durable ideas tend to sit at an odd intersection: something the founders want, can actually build, and that few other people think is worth doing. The first two conditions are obvious. The third is quiet, and it’s where the edge lives. If an idea is plainly worth doing, it’s already crowded. The opening tends to be exactly where capable people have looked and shrugged.
Apple is the clean example. Wozniak built the computer he himself wanted, not a product he’d sized up in a market study, and for a while almost no one else could see why anyone would want a computer at home. The same pattern runs through Google and Facebook, which started as things the founders built for themselves and that most serious people initially waved off as toys. Building for yourself keeps you honest, because you can tell immediately when the thing is bad, and it points you at the unglamorous problems the ambitious crowd has already dismissed.
Underneath all of it is a habit I try to keep: work slightly outside your depth. Not recklessly past it, just slightly outside, in the zone where you’re stretched enough to grow but not so far that you go under. That discomfort isn’t a reason to hold back. It’s usually the sign you’re in roughly the right place.
Dare to act, dare to take on the challenge, dare to own the outcome. None of it happens on its own. You make it happen.
Referenced: Paul Graham, Do Things That Don’t Scale and How to Get Startup Ideas.