FirstSortedValue and Aggr Chart Functions in Straight Table - Dimensions are ignored?

[Somsok]

Hi all,

I am trying to create a table displaying the Customer with the most sales for each Product: 

Customer	Product	Sales
Cust 1	Prod 1	1
Cust 1	Prod 2	11
Cust 2	Prod 1	2
Cust 2	Prod 2	2
Cust 3	Prod 1	3
Cust 3	Prod 2	3
Cust 4	Prod 1	3

I found the chart function FirstSortedValue to get the top customer for each product.
However, only the first row is displayed and incorrect. I think the problem is the inner agg() function. I also visualized this part of the function for testing:

In my understanding sum(Aggr(sum(Sales), Customer)) should only include sales of Prod 1 in row 1 and sales of Prod 2 in row 2. Can anybody can help me out with this? Why does row 1 include Prod 1 and Prod 2 numbers even though the Product Dimension value is Prod 1?

Tanks for your Help!

[Or]

Dimension: Product

Measure: FirstSortedValue(distinct Customer, -Sales)

At least based on the sample data, I'm not sure why you need that messy aggr() in there.

[Somsok · Accepted solution]

@Or Thank you for the input! 
My Data table is highly simplified. In the real case I will have multiple rows of the same Customer and Product I want to sum up. Your solution works if I want to display the Customer with the highest single order. However, I want to see which Customer has the highest order total.

I found a working solution for me:
FirstSortedValue(distinct Customer, -Aggr(sum(Sales), Customer, Product), 1)
I just had to also include the Product in the Aggr statement.