Say your project aims to have a monkey recite Shakespeare on a pedestal. It is then a mistake to invent the pedestal first and leave till later, the difficult problem of training the monkey to speak.
Always start with that step of your project on which success most depends.
Observations 