[dfgdghdf]

> What if it requires research and new types of engineering?

This is even more reason to do market research before sinking a bunch of cost into R&D!

The two cases where you can spend big on R&D without much in the way of market validation are:

1. The product is obviously useful, e.g. self-driving trucks

2. You are going for a patent / aqui-hire play

Point (1) is particularly dangerous when you get attached to your product idea.