The first thing is that you should have a general idea where your play is, whether on the queenside, center, or kingside. Your plan as to where your place your pieces, and how you open that part of the board is going to be based on the opening pawn structure.
If it is later in the game, you want to try to figure out where the opponent is weakest. From there, you want to figure out which of your pieces you can aim at that weakness. This is called "planning".
From there, you have to analyze the best order of moves to move your pieces to attack that weakness, while taking into account your own weaknesses, knowing that your opponent will be trying to do the same to you. Sometimes, you have to just wait and defend if you are much worse.
A lot of this is intuitive based on your level of play, but from there, you just have to calculate variations.
Lastly, as I have gotten older, and tend to miss obvious things more, I like to look at every possible move for both sides, no matter how absurd a move seems. I am not analyzing them, but sometimes, it is just that simple consideration of a move that makes me see a tactic I might have missed, or a mistake I am about to make. I am only talking about spending seconds total looking at all the moves.
Mark Dvoretsky and Artur Yusupov, and I think I have seen other strong GMs say it, say that most calculation errors happen in the first couple of moves.