Simple things, strategical policies

Things are pretty straight. Once you know that you have to start for the quest, making decision for the starting point is really important.
Things are pretty similar in context of that time, when you are creating something, making a product, delivering a code, designing an art piece, solving a problem.
Once you start, things we never thought of, problems we never imagined of, come along the way and you are too busy to rethink about how you started.
Though being strategical is not enough, means if you are putting your efforts just in 'how to start' then you will never care to start, because once we start problems are indispensable. However a strategy about starting and how you are going to tackle them along the journey is important.
Backtracking and re considering your starting point is not a sign of cowardice in this context, because you have to deliver something that is really useful.
Problem solving part is an engineering process, we think in terms 'how' from zero to the top at the stack of solved stages, but what and where and how to start is as important as the problem and as its targeted audience is.