The five rules I would follow to find a startup idea in 2025

I started neural frames at the end of 2022, after a year (or ten) of soul-searching and a couple of deep dives into different, mostly VC-backed startup ideas.

I ended up with something I could hack together myself, bootstrap, and ultimately turn into a profitable business - while having the greatest joy of my life doing so.

Building startups was already in a great place in 2022, with all the services available to make life easier (Stripe, analytics, SEO tools, social media to promote your stuff), but I think it’s gotten significantly better since then. In 2025, we’re in the golden years of doing your own thing: it’s never been easier to build something, and with the rapid advancement of technology, there’s a giant sea of opportunities that opened up, waiting to be captured by a startup.

If you’re trying now, you’ll have a few shots at landing the right thing. If I were looking to found something today, here’s how I’d approach it:

1. Pick a problem you know

You can only come up with meaningful ideas in spaces you have experience in. You can gain that experience by talking to tens or hundreds of people in a particular field and understanding their problems. Or, in the more natural approach, you pick a problem you’ve faced in your own life.

In 2025, literally any task that used to require manual labor can be reimagined with AI: transferring data from one place to another, automated content creation, SEO bots, invoice payment, research & analysis tasks - the list goes on.

If it’s a problem you’ve dealt with personally, there’s a good chance it interests you, too. That matters. If your project succeeds, you’ll spend an ungodly amount of hours on it. And while we shouldn’t overestimate “intrinsic motivation” (I think more often than not, motivation just comes as you get better at something), it still shouldn’t be in a field you don’t care about at all. You’ll need to innovate in this space and generate new ideas - much easier if you’re actually interested.

Ask yourself: Does my experience so far put me in a good position to contribute to this space?

For instance, I’d been making music for years and felt the pain of needing music videos. I also played with synthesizers a lot and loved tweaking knobs and seeing what happened. Both fed into neural frames. Steve Jobs called this “connecting the dots.” But don’t overthink it - these connections often happen naturally and are always easier to tell in hindsight than beforehand.

2. Pick a good problem

The best problems are hard problems for a few people. This is much better than a kinda irrelevant problem for many people.

Typical venture-capital-backed startups go after billion-dollar markets. If they don’t see that potential, they won’t bother. That’s an opportunity for builders who don’t want VC. Focus on a painful thing for a small group.

A big help is analyzing search traffic for certain keywords (using something like Ahrefs). If you’ve got an idea, find keywords people might use when looking for solutions to that problem. With neural frames, that was “AI animation generator.” When I first checked, it had 1,000+ monthly searches in the U.S. alone and low competition. That told me there was real interest. If you find something similar for your problem, you’re onto a winner. Try to rank for that keyword right away.

3. Try to sell as soon as possible

You’re not running a charity. You’re building a business. A business means making money. There’s nothing dirty about it. Without revenue, you likely can’t keep the lights on. Without revenue, you won’t be able to continue solving this problem for your customers. So it’s a win/win if customers pay you to solve their problem.

Charging people early on is also a signal. It shows whether anyone is actually willing to spend money on that specific solution. If enough people will, you have a winner.

In the indiehacking community it got quite popular to even ask for money before building the product. If you can pull that off, more power to you.

4. Think problem first, technology second

I often hear techies say they want to build something “with XYZ technology.” If that’s you, sure - go for it as a side project while working for The Man.

But we’re grown-ups who want to build a business. Customers don’t care about the tech behind what you do; they only care if you solve their problem.

If you stay focused on the problem and use whatever tech is needed, you’ll be so busy you won’t have time to chase every shiny new tool. Putting technology first and problem second will always lose to the opposite. Don’t lose.

Where technology is important is in questions such as “What did technology recently enable that wasn’t possible before?”

5. Pick something you don’t need external money for

Controversial take: Raising money is for people who actually need it. Sure, you can dream of being the next Mark Zuckerberg and building a billion-dollar empire. If that’s your path, awesome, I’ll root for you.

But for most builders in 2025, raising VC can slow you down. Money comes with strings, expectations, and a push to tackle a “big enough” market. VC math only works if you have a serious chance of becoming a billion-dollar company. I’ve seen so many companies shut down, that were doing perfectly fine, but their business model or market didn’t make sense in a venture-backed context.

The good news is, building digital products in 2025 is easier than ever. Startup costs are tiny:

• AI-assisted IDEs let you build an MVP ridiculously fast

• Hosting the MVP costs next to nothing (Vercel, Heroku, Hetzner)

• Stripe makes receiving payments so easy, it’ll be twice the fun to earn money.

• AI tools massively simplify your marketing (especially on social media)

All you really need at first is your own time, focus - basically FAFO.

Think of it this way: If you can’t build a basic MVP with minimal resources, maybe it’s not the right idea - or the right timing. Bootstrapping also means you don’t need anyone’s permission to pivot, slow down, or change direction. That’s invaluable in the early days when you’re still finding your footing.

That’s it for the week. Subscribe if you want to follow my journey. And as always, reply to this mail to tell me what I should write about next 🙂 

AI Video of the Week

I’m a big fan of MINDSET OFF, a German band which creates their psychedelic visuals on neural frames. They deliberately use very abstract settings, and I think they knock it off the park with those.