The Role of Dimensions in the OLAP Cube

I run across this concept of these dimensions existing at different levels of granularity. For example, time could be year, month, day. Could you explain that role of these different granularities in the OLAP cube? Right. So with dimensions oftentimes there's sort of like a hierarchy that can exist within dimension and this issort of like a natural way to classify data. And the way you sort of structure this data is directly correlated to the types of queries you want to make with this data.