Lesson 2 of 6 · 6 min
Start with a problem you already know
You don't need a genius idea. You need a real problem. Learn how to find business ideas hiding in your own daily life, your job, and your hobbies.
You do not need to invent the next big thing. The best first business ideas usually come from a problem you already have. When you build for yourself, you understand the user perfectly because you are the user.
Many beginners think they need a brand new idea that nobody has ever thought of. That is not true. You just need to solve a problem that annoys you. If it annoys you, it probably annoys thousands of other people too. Your own life is the best place to start looking.

Case study
The guitar student who fixed his own practice
What they did first
- He was learning guitar and struggled to get the right sound.
- He tried using ChatGPT to find guitar tones, but it was not accurate.
- He asked other guitarists on Reddit and Instagram if they had the same problem.
- He realized there was no good tool for this, so he built one.
Take this with you
Kyan did not look for a random market. He looked at his own hobby and fixed a frustration he felt every day.
“I realized I was using ChatGPT at the time to figure out how to get guitar tones, but it wasn't really all that accurate.”
This pattern repeats everywhere. Alex Finn was tracking his tweets in a spreadsheet every night. It was boring manual work. So he built software to do it for him. David Bressler was tired of answering the same spreadsheet questions at work. He built a tool to answer them automatically. They did not guess what people wanted. They built what they needed.
$25K/mo
Reported monthly revenue
$226K/mo
Reported monthly revenue
2 days
Days to first version
How to find your idea
Step 1: List your daily frustrations
Do this nowWrite down three things that annoyed you this week. Did a tool break? Did you waste time on a boring task? Did you wish an app existed?
Step 2: Check your job and hobbies
Look for patternsLook at your work tasks and your weekend hobbies. Where do you see people struggling? Where do you see slow, manual work?
Step 3: Ask what people ask you
Listen to requestsWhat do friends or colleagues ask you to help with? If people ask you for help often, that is a sign of a painful problem.
Step 4: Check if you would pay
Be honestAsk yourself: would I pay money to make this problem go away? If yes, that is a strong signal. If no, keep looking.
“The worst case scenario if you build products that solve your own challenges is that you just make your life easier.”
Copying a trend
- You guess what users want
- You do not understand the pain
- You struggle to explain the value
- You give up when it gets hard
Solving your own problem
- You know the pain deeply
- You know exactly what to build
- You can explain it simply
- You are motivated to finish
Your turn
10 minThe Frustration List
Take a piece of paper or open a note on your phone. Write down three things that frustrated you this week. They can be small. Next to each one, write: 'Would I pay ₹500 to fix this?' If the answer is yes, circle it. You now have a list of potential business ideas.
What you'll have at the end
A list of 3 frustrations with a yes/no on paying to fix them.
Key takeaways
- Start with a problem you have yourself.
- If you would pay to fix it, others probably will too.
- Your job and hobbies are full of hidden ideas.
- You do not need a new idea, just a real one.
Where this comes from


I Built an App with Cursor and Made $100K on Launch Day
Alex Finn · Creator Buddy
Watch on SuperBuilder
From Full-Time Job to $1M AI App in Just Over 1 Year (No Code)
David Bressler · Formula Bot
Watch on SuperBuilder

I Built a $10K/Month App from My Mom's Basement
Jack Friks · Curiosity Quench and PostBridge
Watch on SuperBuilder
Builder Copilot
Knows this lesson · Start with a problem you already know
Hi! I've read the whole Start with a problem you already know lesson.
Ask how it applies to you: your budget, skills, city or idea. I'll point you to the parts of the lesson that matter.
Try asking
Your chats are saved, so Copilot remembers what you tell it.