The Metagame

Notes on Paul Graham's "How to Get Startup Ideas"

I wrote a number of essays in 2019 but didn’t publish or complete most of them. This is one of the ones I did finish.

If you’re an entrepreneur considering multiple ideas for a new startup, how do you decide which to pursue? If you were looking to establish a heuristic for this decision, you’d likely have to start by defining what makes a startup idea good. 

Organic ideas 

In “How to get startup ideas”, Paul Graham says that the best startup ideas are “organic”. 

At YC we call ideas that grow naturally out of the founders' own experiences "organic" startup ideas. The most successful startups almost all begin this way.

One way to develop organic ideas is to fix a problem or a pain-point that you see in your life. Take for instance the case of Drew Houston, who built Dropbox when he noticed how frustrating and inconvenient it was to depend on USB sticks to move files between computers. That personal frustration inspired him to build a solution: Dropbox. Seeing pain, whether you experience it first-hand, or you notice it in someone else can be the first step towards building a valuable company. 

Solving a personal pain point is probably the simplest formula for identifying a startup idea. But not all organic ideas are inspired by pain. The best way to come up with “organic” ideas, PG says, is to “live in the future and build what’s missing”. 

What does it mean to live in the future?

“The future is already here — it's just not very evenly distributed.”

Living in the future is to live at the vanguard, to be immersed in what is not yet evenly distributed.

Undoubtedly, the future will solve our present pain points. So if you can see a present pain-point, building a solution is one way to bring us to the future. But the future is also greedy: things that we accept as fine today might be seen as deficient in the future. In fact, because it is not the norm, the future may seem weird. 

PG doesn’t explicitly make a distinction between ideas that are based off of customer pain and ideas that are simply futuristic. Axiomatically, the future entails them both.

But there is a difference between the two: between solving problems for the world as it is and solving problems for the world as it will be. 

Although PG never makes the association, I think it’s helpful to overlay the notion of market risk. If you’re solving a problem that doesn’t require you “to live in the future”, that’s an indication that you’re operating with less market risk. It’s easier when you don’t have to live in the future, when you don’t have to make the bet that the hobbyists or the wierdos presage what will one day become mainstream. In any moment, it’s exceptionally difficult to distinguish between those who are living in the future and those who are just, for lack of a better word, niche.

Inorganic ideas

If you’re in a position where you absolutely need to find an idea, PG recommends a set of tactics that are workable, but categorically “inorganic”. These ideas are discovered indirectly. 

PG lists out a few heuristics:

  1. What are things you used to need (as in, what did you need at your former job?)

  2. Can you turn off your schlep filter?

  3. Garbage-collect broken companies

  4. What customer segment is underappreciated by the big companies?

  5. Is there a wave that you can ride?

My favorite of these is garbage collecting broken companies, although I prefer calling these companies Assassins. They’re marking companies for the kill.

The advantage of these ideas is that they eliminate a great deal of market risk. If you’re going after a company that is already successful, you know (or have reason to believe) that people want what they sell. If it’s the case that most startup ideas are bad because they don’t adequately identify a problem worth solving (inverted: a product people want), competing with an already successful company is one way to ensure that you’re building something that resembles something that people want.  


The most challenging part about finding or even choosing to pursue an organic idea is that you have no control over how long it may take to find. And if you’re impatient, you may pick the wrong one. Hauntingly, it’s impossible to know if the cards you’ve been dealt are the best cards the dealer’s going to send your way. In the early days you can’t really know if your current idea is THE idea or the prelude to THE idea. 

Give yourself some time. You have a lot of control over the rate at which you turn yours into a prepared mind, but you have less control over the stimuli that spark ideas when they hit it. If Bill Gates and Paul Allen had constrained themselves to come up with a startup idea in one month, what if they'd chosen a month before the Altair appeared? They probably would have worked on a less promising idea. Drew Houston did work on a less promising idea before Dropbox: an SAT prep startup. But Dropbox was a much better idea, both in the absolute sense and also as a match for his skills.

If you’re the sort of person who is determined to start a startup, it helps to be mercenary rather than missionary. (I actually hate the missionary/mercenary framing. I think the world is more complex than dividing it between those motivated by love and those motivated by money. As far as I’m concerned, technology is one big wheel we’re all pushing together. Some of us push it more meaningfully than others, but for the most part, despite the notable exceptions: this is noble work.) 

I’m nearly tempted to say, if you want to start a startup because you want to pursue the adventure of starting a company, start an inorganic idea, and be honest with yourself about that. Finding a startup idea that is going to be profitable is hard enough. If you increase the complexity of your search by tacking on the problem of trying to find something that you’re truly passionate about, that’s quite a bit harder.

Harder does not mean worse. But if you set out on this hardest of hard paths--deliberately seeking an idea that you’ll be passionate about--you have to accept that this takes patience and the world’s going to continue turning while you “tinker” and “explore”. Maybe, if you’re adept at the pattern matching game of identifying inorganic ideas, the thing you need to unlock to get started is to hack your psychology: what is it going to take for you to get passionate about what you’re building?

Can you force love for an idea? Can you push through the muck for an idea you don’t love?

I said recently to a founder who wasn’t sure if they loved their startup idea: your work doesn’t need to be sacred for it to matter. The sport of building something and beating competitors and working with great people is sufficient reward. Anything more, and your cup runneth over.