Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.

Reply
nathov
Regular Visitor

Intersection - Retention Rate

I need to create a graph to show customer retention rate.

I'll give an example of the rational behind this KPI:

If I want the retention rate for April this year, I need to look at all the different customers with purchases in the last 3 months (April, March and February) and compare with the customers before that period that would be January. So if 500 customers bought between Apr / Mar / Feb and of those 500, 420 had also bought in January, the retention rate is 420/500 = 84%.

Then I always need to compare the intersection between customers with purchases in the last 3 months with the previous month.

 

I had created 2 measures:

 

Retention Rate 1 =
VAR varHighMonthSelected90days = SELECTEDVALUE(Date[FirstdayofMonth])
VAR varLowMonthSelected90days = DATEADD(Date[FirstdayofMonth];-2;MONTH)
VAR varPreviousMonthSelected90days = DATEADD(Date[FirstdayofMonth];-3;MONTH)

--Returns a table with a column with all customers who purchased within 90 days--
VAR varActiveCustomers90days =
CALCULATETABLE(VALUES('Account'[AccountNumber]);
'Date'[FirstdayofMonth]>=varLowMonthSelected90days &&
'Date'[FirstdayofMonth]<=varHighMonthSelected90days
)

--Returns a table with a column with all customers who bought in the period of 1 month prior to 90 days--
VAR varCustomersActiveMonthPreviousPeriod90Days=
CALCULATETABLE(VALUES('Account'[AccountNumber]);
'Date'[FirstdayofMonth]=varPreviousMonthSelected90days
)
--Returns the Intersection between the two tables--
VAR Intersection = INTERSECT(varActiveCustomers90days;varCustomersActiveMonthPreviousPeriod90Days)

RETURN

COUNTROWS(Intersection)


Retention Rate 2 =
VAR varPreviousMonthSelected90days = DATEADD(Date[FirstdayofMonth];-3;MONTH)

--Returns a table with a column with all customers who bought in the period of 1 month prior to 90 days--
VAR varCustomersActiveMonthPreviousPeriod90Days=
CALCULATETABLE(VALUES('Account'[AccountNumber]);
'Date'[FirstdayofMonth]=varPreviousMonthSelected90days
)

RETURN

DIVIDE([Retention Rate 1];COUNTROWS(varCustomersActiveMonthPreviousPeriod90Days))

 

However, this formula is not working for the beginning of this year (jan/feb/mar). That way, I can't create a chart either, just an indicator according to the month I select. Can anyone help me please?

1 ACCEPTED SOLUTION
Anonymous
Not applicable

Here's the measure:

 

 

// Assumptions:
// 1. There's a dimension that stores Customers.
// 2. There's a fact table, Sales, that joins on
// 	  CustomerId to Customers and Date to Calendar.
// 3. There's a dimension Calendar which is a proper
//	  calendar in the model (marked as such).

[Retention Rate] =
var __lastVisibleDate = LASTDATE( 'Calendar'[Date] )
var __shouldCalculate =
	// You can calculate the ratio only when
	// you can go back in time 4 months from
	// __lastVisibleDate.
	NOT(
		ISBLANK( NEXTDAY( DATEADD(__lastVisibleDate, -4, month ) ) )
		&&
		ISBLANK( DATEADD( NEXTDAY( __lastVisibleDate ), -4, month ) )
	)
var __result =
	if( __shouldCalculate, 
	
		var __custWithPurchWithinLast3Months =
			CALCULATETABLE(
				VALUES( Sales[CustomerID] ),
				DATESINPERIOD(
					'Calendar'[Date],
					__lastVisibleDate,
					-3,
					MONTH
				)
			)
		var __lastVisibleDateMinus3Months =
			dateadd( __lastVisibleDate, -3, month )
		var __custWithPurch1MonthBeforeLast3Months =
			CALCULATETABLE(
				VALUES( Sales[CustomerID] ),
				DATESINPERIOD(
					'Calendar'[Date],
					__lastVisibleDateMinus3Months,
					-1,
					MONTH
				)
			)
		var __custThatBoughtInBothPeriods =
			INTERSECT(
				__custWithPurchWithinLast3Months,
				__custWithPurch1MonthBeforeLast3Months
			)
		var __ratio =
			DIVIDE(
				countrows( __custThatBoughtInBothPeriods ),
				COUNTROWS( __custWithPurchWithinLast3Months )
			)			
		return
			__ratio
	)
RETURN
	__result

 

 

 

Please bear in mind that this works for ANY SELECTED PERIOD OF TIME, not only for months.

 

Best

D

View solution in original post

4 REPLIES 4
Anonymous
Not applicable

Here's the measure:

 

 

// Assumptions:
// 1. There's a dimension that stores Customers.
// 2. There's a fact table, Sales, that joins on
// 	  CustomerId to Customers and Date to Calendar.
// 3. There's a dimension Calendar which is a proper
//	  calendar in the model (marked as such).

[Retention Rate] =
var __lastVisibleDate = LASTDATE( 'Calendar'[Date] )
var __shouldCalculate =
	// You can calculate the ratio only when
	// you can go back in time 4 months from
	// __lastVisibleDate.
	NOT(
		ISBLANK( NEXTDAY( DATEADD(__lastVisibleDate, -4, month ) ) )
		&&
		ISBLANK( DATEADD( NEXTDAY( __lastVisibleDate ), -4, month ) )
	)
var __result =
	if( __shouldCalculate, 
	
		var __custWithPurchWithinLast3Months =
			CALCULATETABLE(
				VALUES( Sales[CustomerID] ),
				DATESINPERIOD(
					'Calendar'[Date],
					__lastVisibleDate,
					-3,
					MONTH
				)
			)
		var __lastVisibleDateMinus3Months =
			dateadd( __lastVisibleDate, -3, month )
		var __custWithPurch1MonthBeforeLast3Months =
			CALCULATETABLE(
				VALUES( Sales[CustomerID] ),
				DATESINPERIOD(
					'Calendar'[Date],
					__lastVisibleDateMinus3Months,
					-1,
					MONTH
				)
			)
		var __custThatBoughtInBothPeriods =
			INTERSECT(
				__custWithPurchWithinLast3Months,
				__custWithPurch1MonthBeforeLast3Months
			)
		var __ratio =
			DIVIDE(
				countrows( __custThatBoughtInBothPeriods ),
				COUNTROWS( __custWithPurchWithinLast3Months )
			)			
		return
			__ratio
	)
RETURN
	__result

 

 

 

Please bear in mind that this works for ANY SELECTED PERIOD OF TIME, not only for months.

 

Best

D

@Anonymous thank you very much for your help!!!!

mahoneypat
Employee
Employee

Assuming you have Oct/Nov/Dec data for this measure to work for Jan/Feb/Mar, I am not sure why you need the Retention Rate 2 DAX.  It looks like you have all you need in the Retention Rate 1 part.  Would this work if you used it in your Return?

 

Return

Divide(Countrows(Intersection); Countrows(varActiveCustomers90days))

 

Regards,

Pat





Did I answer your question? Mark my post as a solution! Kudos are also appreciated!

To learn more about Power BI, follow me on Twitter or subscribe on YouTube.


@mahoneypa HoosierBI on YouTube


@mahoneypat thank you, you're right! I can have all in only one formula but it still doesn't working:

 

Retention Rate =
VAR varHighMonthSelected90days = SELECTEDVALUE(Date[FirstdayofMonth])
VAR varLowMonthSelected90days = DATEADD(Date[FirstdayofMonth];-2;MONTH)
VAR varPreviousMonthSelected90days = DATEADD(Date[FirstdayofMonth];-3;MONTH)

--Returns a table with a column with all customers who purchased within 90 days--
VAR varActiveCustomers90days =
CALCULATETABLE(VALUES('Account'[AccountNumber]);
'Date'[FirstdayofMonth]>=varLowMonthSelected90days &&
'Date'[FirstdayofMonth]<=varHighMonthSelected90days
)

--Returns a table with a column with all customers who bought in the period of 1 month prior to 90 days--
VAR varCustomersActiveMonthPreviousPeriod90Days=
CALCULATETABLE(VALUES('Account'[AccountNumber]);
'Date'[FirstdayofMonth]=varPreviousMonthSelected90days
)
--Returns the Intersection between the two tables--
VAR Intersection = INTERSECT(varActiveCustomers90days;varCustomersActiveMonthPreviousPeriod90Days)

RETURN

DIVIDE(COUNTROWS(Intersection);COUNTROWS(varCustomersActiveMonthPreviousPeriod90Days))

 

This formula still doesn't work for the beggining of this year (jan/feb/mar), assuming that I have 2019 data.

 

@v-yuta-msft can you please try to help me? You've already helped me with something similar befora. Thank you in advance.

Helpful resources

Announcements
Microsoft Fabric Learn Together

Microsoft Fabric Learn Together

Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City

PBI_APRIL_CAROUSEL1

Power BI Monthly Update - April 2024

Check out the April 2024 Power BI update to learn about new features.

April Fabric Community Update

Fabric Community Update - April 2024

Find out what's new and trending in the Fabric Community.

Top Solution Authors