How (NOT) to find a startup idea ❌💡

You want to build a startup, right? No, you want to solve a problem. Jump in here to learn from my mistakes of what NOT to build.

You want to learn how to find a startup idea. I hope you've already done the "mental & financial preparation" lesson first and are ready.

You're wondering what to work on, whether the problem you want to solve is big enough (and... real), whether you'll be able to get your first customer, and whether you'll be able to make a living from it.

So… what should I learn?

To start, there's no better reading than Paul Graham's famous essay:

Right at the beginning, he addresses the most important issue you need to keep in mind throughout the whole process:

"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.
The very best startup ideas tend to have three things in common: they're something the founders themselves want, that they themselves can build, and that few others realize are worth doing."

Even though I knew this, I’ve read this, I was aware of this…I found myself in a trap of building something I didn't really need, couldn't build and was done by multiple vendors. Don't be like me.

I had the luxury of being able to ask experienced entrepreneurs and bounce my stupid ideas off them and kill them early on. If you have such people, good for you, if not: take a moment to watch YC’s content (i.e. from their Startup School), as these people are worth trusting.

Another great piece of advice comes from the evergreen "Zero to One" by Peter Thiel.

He argues that if you copy what's already existing you haven't learned from the greatest. Building competition to what already exists puts you in the field of low margins, thus lowering the potential ROI for investors.

How I understood his book is that startups should question the status quo and rethink business from the ground up. Only by moving towards a monopoly in a given field/niche can you afford "well-being": you have high margins, you're able to survive and finance additional things, be ethical, take care of employees, etc. - if you're in a low margin, you're always fighting for survival.

Don't fall into a trap of "megalomania" though, his advice on building a monopoly is as follows:

1. Start from a small market - it's much easier to dominate a small niche. If you think the market is too big, it probably is. It doesn't mean the market doesn't exist though - a perfect place for a startup is a concentrated number of underserved clients.

2. Don't "DISRUPT" - it's a fancy word, but when our company is described only in terms of differences with existing ones, it means it has nothing special to offer.

3. Last will stay here - you don't have to be first, you just need to do best what is a real problem → make the last improvement that will last for a long time.

It's not bad to re-think what already exists, but make sure that:

• You're really an expert in that field and people will trust you, as why the hell should they go with you instead of the existing market leader?

• The proposal you're making is going to provide "10x". 10x better experience, 10x faster time to market, 10x lower costs, you name it - it just can't be "I built X but slightly better"

You don't want to do things that many people "kind of want", you want to do something that few people "really want".

So… what should I avoid?

Most importantly: avoid TAR PITs. YC will tell you, that every batch there’s thousands of applications that looks amazing and original, but it ends up with lots of people try and don’t succeed and can not pivot from.

I forgot about this falling into the trap of "I want to start" and burned myself. Again, don't be like me.

1. I started with an idea of software house for a specific niche. I figured out a business plan, who we need to hire and even a culture 😆
   I’ve heard: can you answer “what are we selling” and “for whom are we selling this”?
   Well, I couldn’t. Pivot.

2. I decided it has to be a product & I fell into the trap of "let's do AI”. I figured out what’d be the cost of an infrastructure, how much we need to before it’s launched.
   I’ve heard: can you answer which problem you’re solving?
   Well, I couldn’t. Pivot.

3. Then, I thought about how my expertise in crafting data strategy can help products, how they can utilize dashboards, etc (I will spare you the details).
   I’ve heard: “it’s pulled out of your ass”.
   Well, it was. Pivot.
   

It took me more than 3 months to realize, that . It was exactly when “The Truth About Building AI Startups Today” came out. I was laughing & crying at the same time - everything they said on being wrong, was me for the last 3 moths.

This was the moment that made me return to eCommerce, as it turns out it’s the only field I can create value from day 1. Still, there are challenges about “problem might be synthetic”, “what’s the moat”, and so on - yet all the work is more realistic and close to the ground. I know whom to ask & how to validate.

Fun thing: once I return to the roots and expertise, AI appeared by itself as a solution to the problem we’re working on. AI itself isn’t a problem to work on. 

Ok smart-ass, how do I start?

I thought I knew what I was doing, but the truth was way off. Frankly speaking, you won’t learn those until you start, so I’d say… start anywhere?

• Find a problem first that won't be synthetic - remember that the biggest mistake is solving a problem that no one has. Additionally, make sure that this problem is really urgent and painful.

• If you're a business founder FIND A CTO (there’s no excuse for that!). You don't want to dream - you need to have a capacity to push the startup from idea to implementation.

  • (oh, and learn some low-code for MVP, you’ll thank me later!)

• Remember that it’s easier to kill manifesto, landing page, designs in Figma that you can build almost in no-time, than to spend months on building products that makes no sense.

  • (oh, and learn Figma prototyping, you’ll thank me later!)

• See request for startups: just don't fall into the trap that since it's roughly described there, you now have an excuse from looking for a problem. Don't start with "startup ideas", start with "problems that exist".

Next time I’ll write about product discovery without a product - it’s a fun thing to do, if you’re ready to be rejected 9/10 times and have an emotional rollercoaster along the way 🥹
But… as Paul Graham said:

Live in the future, then build what's missing.
...
Live in the future and build what seems interesting.

And, well, I'd argue that the future is here, it's just not evenly distributed. So our job as founders is to get dirty and figure out which direction the world is likely going. Then make it happen.