How to Generate Startup Ideas That Are Actually Worth Building

Most founders come to me having already picked an idea. The question they're asking is "how do I validate it?" But the better question — the one that would save years of time — is "how do I know this was the right idea to pick in the first place?"

The best startups start with a problem, not a solution. Here's how to find those problems.

Where good startup ideas actually come from

Domain expertise: you spent 5 years in a specific industry and you know exactly what's broken. This is the most reliable source. You understand the problem deeply, you have relationships in the market, and you have credibility with customers and investors.

Recent transitions: the world changed, but products didn't catch up. Regulation changed. A technology became cheap. A behaviour shifted. The window for a new solution opened. Founders who notice these transitions early have a first-mover advantage.

Frustrated customer experience: you tried to do something and the existing solution was so bad you thought "I could build something better." Be careful here — this often leads to solutions that are better for people exactly like you but not for anyone else.

Second-order problems: you're building in one space and you keep running into the same friction in an adjacent space. That friction is often an underserved market.

The problem-quality filter

Not all problems are equal. Before you commit to an idea, run it through this filter:

Frequency: how often does the target customer experience this problem? Daily pain is much more fundable than annual pain.

Intensity: how much does it cost (in time, money, or risk) when the problem occurs? Low-cost problems don't generate willingness to pay.

Existing solutions: what does the target customer do today? If the answer is "nothing," the problem might not be painful enough. If the answer is "they pay a lot for a mediocre solution," you have a real wedge.

Market size: is there a large enough number of people with this problem at this pain level? A problem that affects 100,000 businesses paying ₹50,000/year is a ₹500 crore market — real. A problem that affects 10,000 people paying ₹500/year is a ₹5 crore market — probably not worth building a venture-scale business on.

Your edge: why are you the right person to solve this? If the answer doesn't differentiate you from 10 other smart people, the market will be contested.

The test that filters out 80% of bad ideas

Find 5 people who are supposed to have this problem. Ask them: "What is the hardest part of [problem area] for you right now?"

If they describe a completely different problem than the one you're planning to solve — that's a signal to rethink.

If they describe your exact problem but say they've solved it already — that's a different signal.

If they describe your problem, get emotional about it, and tell you they've been looking for a better solution — that's the green light to continue.

The 100 startup ideas list as a shortcut

One way to accelerate idea generation is to expose yourself to a large, curated list of problems that have been validated as real and fundable — then filter based on your own expertise and interest. Think of it as a menu of problems, not a menu of solutions.

The best ideas often come from seeing a problem someone else identified and thinking: "I know exactly who has this problem and why they don't have a good solution yet. And I happen to have the background to build it."

That intersection — a real problem, matched to your specific expertise — is where the best startups start.