Rolling customer order tally

This is pretty simple. You have several options.

The problem - display sales for each customer for a rolling 4 weeks.

I'll assume you have a customer dimension, a date dimension, and a sales fact.

Option 1: Add a field to filter on Rolling 4 Weeks

You'll need to add a column to your date dimension. If you can modify the view in SQL Server, do it there. I'll assume a rolling 28 days, but these can be modified to use calendar week boundaries if you want.

If you can't modify SQL Server views, then add the field in Power Query.

If you want to not use Power Query, change your mind. You can do the same in DAX.

--TSQL

CREATE VIEW DimDate AS
SELECT
    ....
    ,Rolling4Weeks =
        CASE
        WHEN [Date] > DATEADD( DAY, -28, GETDATE() )
            AND [Date] <= GETDATE()
        THEN CAST(1 AS BIT)
        ELSE CAST(0 AS BIT)
        END
FROM DateTable



//Power Query

let
  Today = DateTime.Date( DateTime.LocalNow() )
  ,Rolling4 =
    [Date] > Date.AddDays( Today, -28 )
      and [Date] <= Today
in
Rolling4


//DAX

Rolling4Weeks =
DimDate[Date] > TODAY() - 28
    && DimDate[Date] <= TODAY()

Then you can just apply a visual, page, or report level filter on Rolling4Weeks = True. The view, Power Query, or DAX will update every night as your model refreshes. Rolling4Weeks will always be true for the rolling 4 week period.

Alternately, you could make a measure in the Power BI data model:

TotalSales =
SUM( FactSale[SalesAmount] )

Rolling4WeeksTotal =
IF(
    MAX( DimDate[Date] ) > TODAY() - 28
    ,[TotalSales]
)

Your visualization is as simple as a matrix with customers in the rows area, and weeks in the columns area. You either need to use the filter field *OR* the [Rolling4WeeksTotal] measure. No harm in combining them, but no necessity, either.