Forecast Growth with PRODUCTX()

[Anonymous]

I'm trying to forecast based off a variable growth rate built from a "What if" parameter. I've been following this helpful blog Recursive Calculation in Power BI and I've gotten to the final steps and can't seem to get the caculation right

Here is a snapshot of the blog's final excel table based on sales and what I'm trying to mimic. Sales is forcasted out by using the desired growth rates in the final column with ProductX()

I've translated his formulas and got the following results:

All my values seem to line up nicely (last membership caculating correctly, mutiplyby using correct growth rate) but the final forecast fails to caculate and I've been pulling out my hair trying to spot the issue in my DAX. Here are my translated forumlas

LastMembership:

LastMembership = 
IF(
    ISBLANK(CALCULATE(COUNTROWS(FactMembership), DATEADD(DimDate[Date], 1, MONTH))), 
    COUNTROWS((FactMembership)), 
    1
)

MultiplyBy:

MultiplyBy = IF(ISBLANK([LastMembership]), 1+[GrowthRate Value], [LastMembership]) 

Final forecast Calc:

ForecastMembership = 
IF(ISBLANK([TotalMembers]), 
    CALCULATE(
        PRODUCTX(VALUES(DimDate[Month]), [MultiplyBy]), DATESBETWEEN(DimDate[Date], BLANK(), MAX(DimDate[Date]))
    ), 
    [TotalMembers]
)

Can anyone spot where I went wrong?

[Anonymous]

Link to pbix with cleansed data:

https://1drv.ms/u/s!AriLaPnfLC6Fiye7PSQIkN9MXccn?e=b0kbgJ

[v-xuding-msft]

Hi @Anonymous ,

By my test, you need to modify the measure “LastMembership”. The following is my sample you can reference, then to modify yours.

I have three tables. Table DimDate is a calendar table.  There is a many-to-one relationship between table FactMembership and table DimDate.

DimDate = CALENDARAUTO()

Month = MONTH(DimDate[Date])

Note : Create the measure LastMembership using SUM function, not COUNTROWS function.

LastMembership =

IF(

    ISBLANK(CALCULATE(SUM(FactMembership[Members]), DATEADD(DimDate[Date], 1, MONTH))),

    SUM(FactMembership[Members]),

    1

)

GrowthRate Value = CALCULATE(VALUES(ExpectedGrowth[ExpectedGrowth]), LASTDATE(ExpectedGrowth[Month]))

MultiplyBy = IF(ISBLANK([LastMembership]), 1+[GrowthRate Value], [LastMembership])

TotalMembers = SUM(FactMembership[Members])

ForecastMembership =

IF(ISBLANK(FactMembership[TotalMembers]),

    CALCULATE(

        PRODUCTX(VALUES(DimDate[Month]), [MultiplyBy]), DATESBETWEEN(DimDate[Date], BLANK(), MAX(DimDate[Date]))

    ),

    [TotalMembers]

)

Best Regards,

Xue Ding

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

[Anonymous]

Thanks for the in-depth reply! Sadly, that doesn't seem to be the issue. My FactMembership table is a snapshot of my company's membership on a monthly basis where each row is a person and frozen at the end of each month. I can't really sum up the rows but I got around this by adding a caculated column called "members" that just equals 1 and summed that:

LastMembership = 
IF(
    ISBLANK(CALCULATE(SUM(FactMembership[Members]), DATEADD(DimDate[Date], 1, MONTH))), 
    SUM(FactMembership[Members]), 
    1
)

 I've had problems before with time based functions before so I decided to re-create your setup as closely as possible and ended up with this data model using CALENDARAUTO().

Sadly I end up with the same issue with all formulas unchanged except for TotalMembers and LastMembership which I switched to SUM():

Let me know if you can think of anything else. Thanks @v-xuding-msft