An article explaining an implementation of Like for Like (L4L).  Which is a way of analysing how a measure on a group of entities is really changing over time, by removing the influence of changes to the make-up of the group.

    For example.  If your business has many stores, and you can see that overall your sales are going up or down, how do you know if that is due to an increase or decrease in sales at individual stores, or merely because stores have been added and removed from the group.