new lost customers based on current and last year volume

[Anonymous]

What I am trying to implement is something like this. Query to obtain the same is given below.

Segment	Last Period	Current Period	Change	Chagne %	Retained	Retention
Decreasing Customer	150	78	-72	-48%	78	52%
Increasing Customer	88	225	137	156%	88	100%
Lost Customer	74	0	-74	-100%	0	0%
New Customer	0	76	76		0
Total	312	379	67	21%	166	53%

I am trying to get a general idea on how to deal with subqueries and grouping. Checked some methods with "summarize" but can't get it to work. 

created a table with vol_cy and vol_ly grouped on account number and name. But trying to figure out to get the remaining part to work is not leading me anywhere

SELECT

SEGMENT
, COUNT(*) AS ACCOUNTS
, SUM(CY_VOL) AS CY_VOL
, SUM(LY_VOL) AS LY_VOL
, SUM(RETAINED_VOL) AS RETAINED_VOL
, CASE WHEN SUM(LY_VOL) > 0
THEN SUM(FLOAT(RETAINED_VOL))/SUM(FLOAT(LY_VOL))
ELSE 0 END AS RETENTION

FROM(

SELECT
AccountNumber
, AccountName
, CY_VOL
, LY_VOL
, CASE WHEN CY_VOL > 0 AND LY_VOL = 0
THEN 'New'
WHEN CY_VOL = 0 AND LY_VOL > 0
THEN 'Lost'
WHEN CY_VOL >= LY_VOL
THEN 'Increase'
ELSE 'DECREASE' END AS SEGMENT
, CASE WHEN CY_VOL > 0 AND LY_VOL = 0
THEN 0
WHEN CY_VOL = 0 AND LY_VOL > 0
THEN 0
WHEN CY_VOL >= LY_VOL
THEN LY_VOL
ELSE CY_VOL END AS RETAINED_VOL

FROM(

SELECT
AccountNumber
, AccountName
, SUM(CASE WHEN YEAR = 2018 THEN VOLUME ELSE 0 END) AS CY_VOL
, SUM(CASE WHEN YEAR = 2017 THEN VOLUME ELSE 0 END) AS LY_VOL

FROM Table1

WHERE YEAR IN (2017, 2018)

GROUP BY
AccountNumber
, AccountName

)T

)TT

GROUP BY SEGMENT

[v-frfei-msft]

Hi @Anonymous,

Could you please share the pbix or sample data to me, if you don't have confidential data. Please upload the files to One Drive and share the link here.

Regards,

Frank

[Anonymous]

Data is Confidential. What I tried is

I split the query into 2 parts

First summarize() will do the  

SELECT
AccountNumber
, AccountName
, SUM(CASE WHEN YEAR = 2018 THEN VOLUME ELSE 0 END) AS CY_VOL
, SUM(CASE WHEN YEAR = 2017 THEN VOLUME ELSE 0 END) AS LY_VOL

FROM Table1

WHERE YEAR IN (2017, 2018)

GROUP BY
AccountNumber

DAX:

Grouped =

SUMMARIZE('Table1, 

AccountNumber, 

AccountName,

"VOlume CY", CALCULATE(SUM(Table1[Volume]), FILTER(Calendar, Calendar[Year] =  2018])),

"VOlume LY", CALCULATE(SUM(Table1[Volume]), FILTER(Calendar, Calendar[Year] =  2017])),

Year

)

then use calculated columns to obtain SEGMENT and RETAINED_VOL of the below query.

, CASE WHEN CY_VOL > 0 AND LY_VOL = 0
THEN 'New'
WHEN CY_VOL = 0 AND LY_VOL > 0
THEN 'Lost'
WHEN CY_VOL >= LY_VOL
THEN 'Increase'
ELSE 'DECREASE' END AS SEGMENT
, CASE WHEN CY_VOL > 0 AND LY_VOL = 0
THEN 0
WHEN CY_VOL = 0 AND LY_VOL > 0
THEN 0
WHEN CY_VOL >= LY_VOL
THEN LY_VOL
ELSE CY_VOL END AS RETAINED_VOL

Then used SUMMARIZE() to create final table from the result table "Grouped" created using the first SUMMARIZE()

SEGMENT
, COUNT(*) AS ACCOUNTS
, SUM(CY_VOL) AS CY_VOL
, SUM(LY_VOL) AS LY_VOL
, SUM(RETAINED_VOL) AS RETAINED_VOL
, CASE WHEN SUM(LY_VOL) > 0
THEN SUM(FLOAT(RETAINED_VOL))/SUM(FLOAT(LY_VOL))
ELSE 0 END AS RETENTION

Final =SUMMARIZE(Grouped,

Grouped[Segment],

COUNTROWS(Grouped),

Grouped[Volume CY],

Grouped[Volume LY],

Grouped[RetainedVolume],

"Retention", DIVIDE(SUM(Grouped[RetainedVolume),SUM(Grouped[Volume LY]))

)

However, the resulting values I get doesn't match with the result I get from the SQL query. Can someone let me know if the approach is correct or not?