This one requires some foresight. The square to notice is e6. It looks unassuming, but it's the focus of the pressures on the board; it is attacked twice by White and defended twice by Black. As we know, in this sort of position it pays to ask what would happen if those potential exchanges all were made, so that the board became simplified with a last enemy piece left standing on the contested square: what would then be possible? The train of thought is especially important here because the disputed square is on a diagonal with the Black king, suggesting the possibility of a pin once the smoke clears. White has his light squared bishop positioned to move to d5, making the chance for a pin more than hypothetical.

So think through the sequence: White plays Nxe6; Black replies BxN. White plays RxB; Black replies QxR. Now White plays Bxd5, pinning the queen and winning it a move later. (Of course a piece that pins the queen needs protection, and here the bishop gets it from the rook on d1. The knight on d4 would have been moved out of the way two moves earlier. When you imagine using a piece during a sequence, be careful to imagine that the square it used to occupy becomes vacant.)

In a sense this position, like others we recently have seen, just involves upgrading a target: the pawn on e6 eventually is traded up to a queen. But in practice you might not see the possibility of the pin right away here because the target of it looks so unpromising at first. Still, anytime your bishop is one move from being able to land on the enemy king’s diagonal and pin anything there, it's worth asking whether exchanges on that square might make the pin profitable.