2 min read

Don't Build Castles In The Air

Don't Build Castles In The Air

How prepared are you to launch the MVP?

I’m gonna use this blog to take some notes from contents I need to learn right now.


  • Get the project as fast as you can to test the market.
  • The idea must be flexible, because sometimes what you wanna build may not be useful to the actual customer.


  • Don’t fall in love with the product. Focus on the problem. If your problem isn’t solved, change the product. Iterate the solution until it actually solves a problem.


  • Build it in weeks not months.
  • Landing page and spreadsheet.
  • Most startups can start very fast.
  • LIMITED functionalities.
  • Don’t address all of the user’s problems.
  • Appeal to a small set of users.

AirBnB case:

  • Started without map view.
  • Started without payments.
  • They had a Part-time CTO.

Twitch case (Day one):

  • 1 channel (justin)
  • Low res video
  • No videogames


  • No bank deals
  • Few features
  • Founders would integrate it for your. On YOUR OFFICE.

Heavy MVP

  • If insurance, banking, biotech, moonshots, or any strict business areas, you need a more developed MVP.

Press Launch X Get Customer Launch

Press Launches aren’t as important, do you remember Facebook or Twitter press launch? Focus on getting customers.

It is hard to learn from the customers if they don’t have anything to play with.


  • Time box your spec(what if in 3 weeks?)
  • Write your spec
  • Cut your spec
  • The goal is to get anything out in the world, it is easy to get things going once they are in the world. Else it is easier to just delay.
  • Don’t fall in love with the mvp.


  • It is not the user’s job to give you features, that is your job.
  • The user gives you problems.
  • “Is there anyone else with the same problems?”
  • “Why do you want to have blablablablbal feature? What problems do you have that require that? How often do you encounter that problem?”
  • I’m stuck at the mvp and I don’t know how to launch in weeks? STOP IT. It is not special. Change it.
  • If you think it has to be shitty. You don’t spend a lot of time tinkering with it, it is just a test tool.
  • Does it solve the problem I want it to solve?
  • Understand the problem you are trying to solve.
  • If it is in front of the user, you can see if they want it. If it is a problem that they have every day, they come back the next day.
  • How long should the mvp last? Solve the problem first. A, B, C, D. If you are in the pre mvp stage, focus on A. Find and solve the problem.
  • As a founder you are gonna have to juggle multiple things.


  • People start using your product so much, you end up just trying to keep it online. “I’m pretty sure we are gonna die because we have so many users”.
  • What if the problem expands? It is expected. Don’t make the mistake to solve the problem for all users.
  • Be comfortable that you have 2 problems, the vision to the investors and the actual product you have. API that is so difficult to use that we have to install it ourselves for the clients. But we want to process payments for everyone.

That is it. If you want the video I got this from Y Combinator’s YouTube Channel: