Don't start from solutions

The fastest way to build the wrong thing is to start by looking at what already exists. You study the market, note the patterns and inherit the assumptions. What you end up with is a refined version of someone else's answer to someone else's question.

Start from the problem directly in front of you. Not the category it belongs to. Not the prior art. The actual friction in your hands right now.

The interesting solutions live in the gap between the problem as it is and the solution everyone assumed it needed.