Percentage of customers that bought X quantity

[ChPetru]

Hello,

I am trying to make a matrix visual in which I have the Product Quantity (across all account opportunities) on columns and account value on the rows. Values in the matrix should be the Number of Accounts with that product quantity divided to number of customers in the Account Value bracket. Something like below.

I have issues with: Quantity (since it's at opportunity level in my table, an account will appear twice if he has bought the same product twice on different opportunities. I want the product quantity at account level; this calculated column did not help: 

_Quantity Account Level = CALCULATE(sum(Transactional[Quantity]),Accounts[Account ID ]=EARLIER(Transactional[Account ID ]),Transactional[Product]=EARLIER(Transactional[Product]))

Also, i created a measure to show me the percentage of accounts for each quantity, that also ain't working fine since at matrix row level the percentages do not add to 100%:

Percentage of Accounts = DIVIDE(CALCULATE([_Number of Distinct Accounts],ALL(Accounts[_Account Value])),CALCULATE(Accounts[_Number of Distinct Accounts],ALL(SFDC[Quantity]),ALL(SFDC[Product])))

Filters:	Product1
Account Value / Quantity	1	2	3	4	5	6	7	8	x
$0-$1000
$1000-$2000
$2000-$3000
Accounts		Transactional
Account ID		Account ID1	Opportunity1	Product1	Quantity
Account Value		Account ID1	Opportunity2	Product2	Quantity
		Account ID2	Opportunity3	Product3	Quantity
		Account ID2	Opportunity 4	Product 1	Quantity

Not sure if anyone can help defining the two measures to work as intended.

Expected Output:

Filters:	Product1
										Row Total
Account Value / Quantity	1	2	3	4	5	6	7	8	x
$0-$1000	2%	0%	0%	5%	0%	...	..	..	..	100%

[v-rzhou-msft]

Hi @ChPetru,

Your calculation is based on your data model. Please share a sample file with me and show me a screenshot with the result you want. And you can show me more details about your calculate logic. This will make it easier for us to find the solution.

Best Regards,
Rico Zhou

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.