"What should I build?" is the wrong question.
It's the question people ask when they want to be founders more than they want to solve specific problems.
I see this constantly. DMs, Reddit posts, Discord channels—people asking strangers what they should spend months building.
And here's the issue: if you need someone else to tell you what to build, you're already approaching this wrong.
Why the question is backwards
Building something takes months. It's going to be hard. There will be days when nothing works and you want to quit.
The only thing that gets you through that is genuinely caring about the problem.
Not "I think this would be a good market." Not "Someone told me this was a good idea." But: "This problem is annoying me personally and I want it solved."
When you ask "what should I build?" you're asking someone to give you that conviction. But conviction doesn't work that way. You can't borrow it.
The idea generation trap
People think the hard part is finding the right idea. It's not.
The hard part is:
Shipping consistently
Talking to users
Iterating based on feedback
Not giving up when it's hard
All of that requires intrinsic motivation. You need to actually care.
And you can't care about an idea someone else gave you the way you care about a problem you personally experience.
What actually works
Every successful founder I know started with a specific problem:
Stripe: payments for developers were terrible
Figma: design tools were slow and not collaborative
Notion: organizing information was fragmented
Airbnb: founders couldn't afford rent and had extra space
None of them asked "what should I build?" They built solutions to their own problems.
The market validation came later. First came: "this is annoying, I'm going to fix it."
Why people ask anyway
People ask "what should I build?" because:
1. They want to be founders, not solve problems. The title is more appealing than the work.
2. They think there's a "right answer." Some perfect idea that's obviously going to succeed.
3. They're afraid of picking wrong. Better to crowdsource than risk building something that fails.
4. They don't have problems worth solving. Which means they're probably not the right person to build anything yet.
All of these are red flags. If you need external validation to pick an idea, you probably won't have the conviction to execute it.
The market research fallacy
Some people defend this approach: "I'm doing market research to find problems worth solving."
But here's what actually happens:
You research a problem you don't personally have. You convince yourself it's worth solving based on data. You build something. You realize you don't understand the nuances because you're not the user.
Meanwhile, the founder who actually experiences the problem understands all the nuances. They know what really matters. They talk to users who are like them.
You can't research your way into empathy for a problem you don't have.
When asking makes sense
There are legitimate questions:
"I'm building X, should I focus on feature A or B?" (prioritization, not idea)
"Does this problem resonate with you?" (validation, not generation)
"How did you solve X?" (learning, not outsourcing)
But "what should I build from scratch?" is different. That's asking someone to give you the thing that's supposed to drive you.
What to do instead
If you don't know what to build:
1. Build something for yourself. What's annoying you? What tool do you wish existed?
2. Work at a company and notice problems. Real problems, not hypothetical ones.
3. Talk to people in industries you understand. Find pain points in contexts you actually know.
4. Wait until you can't NOT build something. Until the problem bothers you enough that building is the obvious response.
Don't force it. Don't try to find an idea because you decided "now's the time to start a startup."
The best companies come from people who couldn't stop thinking about a problem until they solved it.
The commitment test
Here's the test: if someone told you this idea wouldn't make money, would you still build it?
If yes → you probably actually care
If no → you're optimizing for outcome, not solving a problem
Most successful founders would have built their thing even if they knew it would be small. Because they wanted it to exist.
People asking "what should I build?" are usually optimizing for "what will make me successful?" which is the wrong optimization.
What I learned building Nexus
I didn't ask anyone if I should build Nexus.
I was a solo founder trying to build a company. I needed operational and growth help but couldn't afford employees. Hiring agencies was expensive and they just advised, they didn't execute.
I wanted AI agents that actually did the work, not just gave me suggestions.
So I built it. For me. Then realized other solo founders probably had the same problem.
Did I validate the market first? No. Did I ask people if this was a good idea? No.
I just built the thing I needed and put it in front of other solo founders to see if they needed it too.
That's how this works. You don't crowdsource your conviction. You find it in actual problems you have.
— Arjun