Count Distinct Values over Time

v-chuncz-msftand Greg_DecklerPerhaps there's something wrong with our data or the formula? I'm trying to calculate a running order count for generic months over time. NOTE: This data is for customers with contracts that all have different start dates, so we had to create a generic calendar with years, quarters and months simply numbered as 1,2,3,4, etc.; hence the filtering below and the reason we cannot use built-in time functions.

 
Whether I use:

Order Count = DISTINCTCOUNT('Analytics'[OrderNum])

Or use Count of OrderNum (Distinct) in the line graph visual, the results are the same:

Counts by Month =

Mo.   Order Count

1              39

2              42

3              33

4              23

Counts YTD are then showing =

Mo.   Order Count

1              39

2              77

3              104

4              125

However, if the Monthly counts are correct, then the YTD counts are incorrect (or vice versa - if YTD counts are correct, then monthly are incorrect). For example, if monthly counts calculated above are correct, then YTD, SHOULD be:

Mo.   Order Count

1              39

2              81

3              114

4              137

Here's the formula I'm using for YTD Count:

Count Orders YTD = CALCULATE ((DISTINCTCOUNT('Analytics'[OrderNum])), FILTER (ALL(Champ_Calendar_Lookup), Champ_Calendar_Lookup[Champ_Calendar_Year] = MAX (Champ_Calendar_Lookup[Champ_Calendar_Year]) && Champ_Calendar_Lookup[Champ_Calendar_Month] <= MAX (Champ_Calendar_Lookup[Champ_Calendar_Month] )))

YTD order counts are coming out too low.

Maybe I'm just missing something simple?

Thanks for any help!