Y Combinator has funded more than 5,000 companies since 2005 (YC press page). In a November 2012 essay, How to Get Startup Ideas, its cofounder Paul Graham summed up the approach in two sentences: "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."
Peter Thiel, cofounder of PayPal, starts somewhere else in Zero to One (2014). He argues that great companies are built on a secret, meaning a truth few people share, and that competition makes everyone poorer. His Wall Street Journal essay of September 12, 2014 is titled "Competition Is for Losers".
This guide walks through both methods and how to apply them: where to look, how to filter, and how to test an idea before you give it a year of your life.
What makes a good startup idea
Good startup ideas rarely come from a flash of inspiration. You find them by watching for everyday problems painful enough that someone would pay to make them go away.
Start with a real problem
Before you map your idea onto a Lean Canvas, you need a concrete problem that costs the people who have it. Investors and accelerators like Y Combinator back founders who go after a specific pain point rather than a concept. When you solve a problem people have already tried to fix on their own, you know demand exists before you write a line of code.
Three questions help you filter. Does this problem cost someone time or money every week? Have those people already hacked together a workaround? Would they pay for a solution that actually works? If the answer to all three is yes, the idea is worth digging into.
Cool idea vs. useful idea
An idea that impresses people at a party and an idea that meets a real need don't always overlap. A "cool" idea gets attention for a week. A useful idea, even a boring one, keeps its users.
A social network for pets gets a smile (cool idea). A service that finds you an available vet within twenty minutes of your home, in under ten minutes, solves an emergency pet owners have already been through (useful idea).
Startups born from a personal need
Three Y Combinator companies began with a problem their founders had:
- Airbnb: Brian Chesky and Joe Gebbia couldn't make rent in San Francisco, so they rented out three air mattresses in their living room during a design conference when every hotel was booked;
- Dropbox: Drew Houston kept forgetting his USB drive and wrote the first prototype on a bus ride;
- Slack: the internal chat tool Stewart Butterfield's team built while developing a video game, Glitch, outlived the game.
All three follow the lean startup method: a personal frustration, a first tool built for yourself, then a product for everyone else.
How to spot startup ideas around you
Ideas are all around you, in the things you already do. To see them, write down the friction instead of putting up with it.
Pinpoint friction in everyday life
Grab a notebook, paper or a notes app, and for two weeks log everything that wastes your time: at home, at work, at university, in your neighborhood. A form that takes an hour, a tool you work around, a purchase you keep putting off because you lack information. Our guides for the next steps are in the swanbase resources. After two weeks you'll have around twenty entries. Two or three will keep coming back, and those are the ones to dig into.
Build for yourself and your friends
Ask the people around you what they're missing. What product or service would help them that they can't find today? An hour with five friends who work in different fields from yours surfaces more leads than a week alone at a whiteboard. When you then build for those same people, you get direct feedback without commissioning a market study.
Watch for unmet behaviors
Look at how people use existing products and where they stray from them: the Excel file that replaces the official software, the WhatsApp group that replaces the business tool, the screenshot that replaces the export. Every workaround points to a need the current product doesn't cover. When you note the triggers, motivations, and obstacles that come before each purchase, you find the gap where a new product fits.
Paul Graham's principles for finding an idea
Paul Graham, cofounder of Y Combinator, lays out a three-part method in his 2012 essay.
Notice recurring problems in your own life
Graham writes that you shouldn't try to invent ideas but to notice problems, ideally ones you have yourself. A problem you live with every day has an edge over all the others: you know it exists, you understand the context, and you are your own first user. The most promising ideas often seem obvious, at least to the person who has them.
Build for a small group of people
Graham frames the choice in one sentence: "You can either build something a large number of people want a small amount, or something a small number of people want a large amount. Choose the latter." At launch, you need users with an urgent need for your product, and that group is small. Airbnb started with attendees of a single conference in San Francisco, when the hotels were full.
"Make something people want," YC's motto
Graham offers a test to tell whether you're onto something: "Who wants this so much that they'll use it even when it's a crappy version one made by a two-person startup they've never heard of?" Until you can name those people, you have a theme rather than an idea. According to CB Insights' analysis of 431 venture-backed companies that shut down since 2023, 43% cite lack of product-market fit as the reason they failed (CB Insights, March 2026).
Peter Thiel's strategies for thinking differently
Peter Thiel, cofounder of PayPal and Facebook's first outside investor, comes at the problem from the other end in Zero to One: instead of a need to meet, he looks for a market to create.
Look for a secret
Thiel calls a "secret" an important truth that few people accept. He sees two kinds: secrets about nature, things the physical world hasn't revealed yet, and secrets about people, things people don't know about themselves or keep hidden. The question he asks every founder and every job candidate: "What important truth do very few people agree with you on?" A good answer describes a market others missed because they didn't dare look at it.
Build a monopoly instead of competing
In his Wall Street Journal essay, Thiel argues that "competition is for losers": in a market with no barriers to entry, profits trend toward zero and no company earns enough margin to invest. A monopoly in Thiel's sense is a company so far ahead of its rivals that none of them can catch up. That kind of company can afford research, good salaries, and a long time horizon, while its competitors fight to survive on thin margins.
Think long term from day one
In Zero to One, Thiel writes that most of a tech company's value comes from cash flows ten to fifteen years out, and that being the last to enter a market with the decisive breakthrough pays more than being first. For a founder, this becomes a question to ask from the start: can this idea still dominate its market in ten years, or will it be copied within eighteen months?
How to go from idea to market
Even an excellent idea is worth nothing until someone has tried it. Four steps take you from idea to a product you can sell.
Validate your startup idea with an MVP
To validate a startup idea, start by building a minimum viable product (MVP): the simplest version that lets a user solve their problem. Speed matters more than polish here. A scrappy MVP built in two weeks teaches you more than a product refined for six months. The Lean Canvas helps you write down the assumptions to test before you code.
Test with a small market
Target a specific segment, show them your solution, and watch what they do with it. Your first ten users will tell you what's missing, what gets in the way, and what they'd pay for. Fix things with this group before you expand.
Think about distribution from the start
How will customers find out you exist? Which channels will you use, and what will each customer cost to acquire? A product without a distribution channel is still a side project. From the design stage, think about how one happy user brings in another.
Match the idea to your skills (founder-market fit)
Make sure the project fits what you're good at and what you care about. You'll spend years on it. Experience in the field gives you access to users, an understanding of their constraints, and the staying power to keep going when the first six months produce nothing.
A plan for the next two weeks
Finding a startup idea takes a method of observation rather than a stroke of genius. Whether you follow Paul Graham's approach, starting from a personal problem, or Peter Thiel's, looking for a truth others refuse to see, the starting point is the same: a written list.
This week, open your friction notebook and write down every daily irritation. In two weeks, keep the three entries that come up most often. For each one, call five people who have the same problem and ask how they deal with it today. If one of those three leads still holds up after five calls, build the MVP.
You can also start a company without a set idea. Slack came out of an abandoned video game, and cofounders often meet before they find their subject: our guide to finding a cofounder explains how. If you have an idea and want founders to challenge it, apply to swanbase: long-term support, €0, by application, and swanbase takes equity.
Frequently asked questions
Can you start a startup without an idea?
Yes. Slack grew out of an internal tool built for another project, and Paul Graham advises starting from a problem you've lived rather than an idea. Begin with a friction notebook and five conversations with people who share the same problem.
How do you know if a problem is worth a startup?
Apply Graham's test: who wants this solution badly enough to use a flawed first version made by two strangers? If you can name ten people and they've already hacked together a workaround, the problem deserves an MVP.
Does a startup idea need to be original?
It needs to be useful to a specific group, and originality comes second. According to CB Insights (March 2026), 43% of funded startups that shut down since 2023 cite lack of product-market fit: the product, original or not, never found its users.
What's the difference between a problem in Graham's sense and a secret in Thiel's sense?
Graham starts from a concrete pain you experience and others share. Thiel starts from a conviction most people reject which, if true, opens a market with no competitors. Both converge on the same final test: users who pay.
