I have item level data that I need to analysis at both customer and transaction level.
I need to identify the number of customers in three different groups:
1. The customers just buying the brand in the current selection and their spend.
2. The customers buying the brand in the current selection and other brands as well. Split by the spend on the brand and non-brand sales.
3. The customers not buying the brand in the current selection and their spend.
I have been trying to use set analysis to do this, but can't seem to get my head around the correct formula. Using the possible select seems a necessary step, but I am not sure whether using the exclusion of the possible select would work (e.g brand=e({$}brand) ) or whether something along this lines is better:
