When analyzing data there are two main pitfalls; analyzing too much and analyzing too little. Analyzing too much means that we spend too much cost on analyzing in relation to the value we gain. Analyzing too little means that we "lose" value because we know too little.
Let me give you two examples:
1. Analyzing the relation of sales between all existing articles would probably be to analyze too much.
2. Knowing only the average ticket of all tickets would probably be to analyze too little.
So, how do we actually know that we are analyzing our data at a relevant level? There is no easy answer, and sometimes the only way to find out is to analyze too much and go backwards from there.