Revisiting YC Co-Founder Paul Graham's Essay: How to Get Startup Ideas | Bolt Recommended Reading
Start by solving your own real problems.
"Organic Startup Ideas" is a 2010 essay by Y Combinator founder Paul Graham, in which he shares two primary approaches to finding startup inspiration and how to create genuinely valuable products by starting from user needs. Though written more than a decade ago, the entrepreneurial mindset it describes remains relevant for founders in the GenAI era.
We've organized and translated the essay below; you can read the original by clicking the "read more" link.
The best way to find startup inspiration is to ask yourself: "What do you wish someone would make for you?"
Startup ideas typically fall into two categories: those that emerge naturally from your own life, solving problems you personally face; and those where you observe unmet needs in others and decide to build something for that user group. Apple falls into the first category. Apple was born because Steve Wozniak wanted a computer. Unlike most people who wanted computers but couldn't design them, he could — so he built one. Because many others wanted the same thing, Apple was able to sell enough computers to grow into a company. Incidentally, Apple still follows this principle today; the iPhone, for instance, was the phone Steve Jobs wanted.[1]
My own startup, Viaweb, fell into the second category. We built software for creating online stores. We weren't the target users ourselves, but because my co-founders and I were relatively old when we started the company (I was 30, Robert Morris was 29), we had enough life experience to recognize that users would want this kind of software, so we built it.[2]
While there's no sharp boundary between the two types, the most successful startups seem to hew closer to Apple's model than Viaweb's. When Bill Gates wrote the first BASIC interpreter for the Altair, he was writing it for himself; so were Larry Page and Sergey Brin when they built the first version of Google.
Solving your own problems tends to yield startup ideas with a higher probability of success, especially when you're young. Predicting what others will need requires experience. At Y Combinator, the worst startup ideas we see often come from young teams building things they think other people will want.
So if you want to start a company but don't yet know what to build, I'd encourage you to focus initially on solving your own problems for inspiration. Ask yourself: "What's missing in my daily life? What could be improved?" Sometimes asking this question yields an immediate answer. In retrospect, I suspect Bill Gates found programming the Altair in machine code to be such an obviously bad experience that improving it was a no-brainer.
Other times, you need to step outside your own perspective, because people tend to get habituated to problems and accept them as given. But rest assured, the problems are there. Great ideas are all around us, waiting to be noticed. In 2004, Harvard students were still using paper Facebook directories — archaic, when the whole thing could obviously be moved online.
Similar opportunities exist today. The reason you overlook them is the same reason you might have overlooked the Facebook idea in 2004: ideas that come from solving your own problems usually don't look like startup ideas at first. We know now that Facebook was hugely successful, but in 2004, putting college students' profiles online didn't seem like a particularly compelling startup idea. In fact, it wasn't a startup idea at all initially. Mark Zuckerberg said at a YC dinner this winter that when he wrote the first version of Facebook, he wasn't trying to start a company — it was just a project. Similarly, Wozniak wasn't trying to start a company when he began working on the Apple I. If these people had been focused on starting a company from the outset, Facebook and Apple might not exist.
So if you want to find good startup ideas, I'd suggest focusing more on the ideas themselves and less on the concept of "starting a company." Just fix things that seem broken, regardless of whether the problem seems important enough to build a company around. If you keep following this thread, it's hard not to eventually create something valuable for many people. And when you do that — congratulations, you have a company.[3]
Don't be discouraged if your initial product is dismissed as a "toy." In fact, this may be a good sign. It's probably why others overlooked the idea. The first microcomputers were considered toys; so were the first airplanes and the first cars. These days, when someone comes to us with something users love that might be mocked as a "toy," we're actually more likely to invest.
While young founders may be at a disadvantage in identifying others' needs, they often have an edge in discovering new problems. The reason is that they're on the leading edge of technology, habitually using the newest things. And because everything they use is so new, they're the first to spot the problems within it — and the enormous opportunities those problems represent.
The most important thing in startups is finding unmet needs. If you've discovered a genuine user need and have the capability to solve it, congratulations — you've found a gold mine. Like a real gold mine, you'll still need to work hard to extract the gold. But at least you know where the vein is, and that's the hardest part.
Notes:
[1] This offers a way to predict potential blind spots at Apple: things Steve Jobs didn't use, Apple may not be great at. He didn't seem to be much of a gamer, for instance.
[2] In retrospect, if I were to redo Viaweb, I'd try running an online store myself. That would have helped us understand our users better. I'd encourage anyone starting out to become a user of your own product if at all possible, even if it seems outside your wheelhouse.
[3] One possible exception: competing directly with open source software is difficult. You can build tools for programmers, but you need to find something you can charge for, or it's hard to sustain.
Thanks to Sam Altman, Trevor Blackwell, and Jessica Livingston for reading drafts of this essay.
📮 More Reading
Linear Bolt Bolt is an investment initiative by Linear Capital dedicated to early-stage, global-market-facing AI applications. It upholds Linear's investment philosophy, focusing on projects driven by technological transformation, and aims to help founders find the shortest path to their goals. Whether in speed of action or investment approach, Bolt's commitment is to be lighter, faster, and more flexible. In 2024, Bolt has invested in 11 AI application projects including Final Round, Xinguang, Cathoven, Xbuddy, and Midreal.