[kowdermeister]

Currently I'm in totally "guilty" of #1, but it feels fine and I don't care if it's not the best way to launch something. I want to sell t-shirts online. What kind of research do I need? It's a proven model, people buy t-shirts and other merch online, so the only thing I need is to build it, design it before selling. I just want to build a side project that pays the bills, not going after to outplace major players.

#2 is interesting, I will do this by first selling to friends and talking to them in person to collect enough feedback to do an iteration on the product. I do agree that listening is key.

#3 The ideal situation for you is to have as many signed, legally binding customer contracts as possible

Meh, this is for one specific type of business, there's tons of business models that this doesn't apply to. If the product can collect money from day one, then you won't have the problem of mistaking interest for demand. What's even better is to turn people who are just interested into your advocates.

[soneca]

If you are ordering the production of 1,000 shirts hoping that design will sell, you are guilty of #1.

If you are just designing it, putting on a website and will only actually manufacture the shirts you sell, then the solution to #1 is already built in your business model (assuming you are not spending 6 months of coding with your ecommerce website).