read

I have often been suprised how quickly do people associate failure with something negative. I mean sure if you poured all your life savings into something and it failed tragically, that does sound very sad indeed. But if it wasn’t your last attempt, failures could be such a rich treasure trove of lessons and stepping stones to your path ahead.

In this article I’ll tell you all the mistakes I did while building CtrlB from late 2022 to mid 2024 so that you and I may avoid it the next time.

[Validation]

• Not a good idea to start with a solution and find a problem. Just because you see a company building a solution and it aligns with your “thought process” or bias, doesn’t mean its a great idea. We saw the founder of CtrlStack talking about problems to which we arrived entirely ourselves and it gave us all the confidence we needed to build a similar solution.
• While validating ideas, people would say the idea is good because they want to make you feel good. It is not their job to validate your idea for you. It is your job to get it done. How do you do that? Just remember 2 things. Do not talk about what you’re building too early, talk about their problems. When you’re at stage where you need to talk about your solution, ask about their paying intent to understand if they really want it. Read more about this in the MOM test book. While validating our live debugger we never talked about paying intents.
• Good idea to build for something for which you can track a north star metric. Otherwise it becomes very difficult to show what value your customers get out of you. Our live debugger improved developer productivity but we couldn’t measure it.

[Fund raising]

• Not a good idea to raise money without a validated plan. Fund raising starts an invisible clock on your head because your investors want to see your progress. It is difficult to show progress when you still figuring out what to do. We raised a lot of money - thanks to our IIT tag - too early.
• Until money hits the bank, no deal has been made. We had people who pulled back from a signed term sheet because they got cold feet last minute.
• Raising funds and creating a business are not exactly correlated. In 2023, we were masters of raising funds just on words, but that didn’t mean we had any real business.
• Don’t hire people if you can’t carve out work for them. If you made the mistake of hiring, fire them quickly.

[Co-founder compatibility]

• If you cannot resolve a conflict with your co-founder, no point working together. Always a good idea to sort it out rather than bearing consequences later.
• More than the idea, more than the business, it’s very important that the co-founders are aligned in their thoughts. When I had started thinking about leaving CtrlB in early 2024, that’s when I realized the differences between me and my co-founder. I was in this more for the experience of building something, while he was in it as if this was his life.
• Living with your co-founder doesn’t sound like a great idea, if you don’t understand each others work-like balance.
• Set deadlines for tasks and hold each other accountable if they aren’t met.

[Miscellaneous]

• Once the validation is done, team needs to be razor focussed on what and for who it needs to be built. If there is a lack of razor sharp focus, that means validation is incomplete.
• Leverage as much of open source tech that you can. It is some idiot’s fault who built some tech and decided to make it open source. Most of the code for our live debugger was available open source at different places. Then we used StarRocks as our open source query engine.
• Your employees are not your friends. By extension, the people you work with are not your friends. You will have to fire them if it comes to it.
• Be selfish. Most people wait for things to happen to them because it is difficult to take a strong step. Remember no one is coming to help you. If you see it isn’t working, leave. Don’t stay put to pacify other shareholders.