Power query

I would do it differently.

I did this:

1. Seleccted the first 3 columns, then unpivoted other columns
2. Filtered the "a" out of the Values column
3. Grouped by the first 3 columns and did a CountRows.

I get this:

If you need the original data again, just merge it with the source step.

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WqlDSUUoE4lQoDcGxOtFKVVVVQDaUhMoUgDFItqICpBNCwCXAkrEA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Column1 = _t, Column2 = _t, Column3 = _t, #"1/1/2021" = _t, #"1/2/2021" = _t, #"1/3/2021" = _t]),
    #"Unpivoted Other Columns" = Table.UnpivotOtherColumns(Source, {"Column1", "Column2", "Column3"}, "Attribute", "Value"),
    #"Filtered Rows" = Table.SelectRows(#"Unpivoted Other Columns", each ([Value] = "p")),
    #"Grouped Rows" = Table.Group(#"Filtered Rows", {"Column1", "Column2", "Column3"}, {{"Count", each Table.RowCount(_), Int64.Type}}),
    #"Merged Queries" = Table.NestedJoin(Source, {"Column1", "Column2", "Column3"}, #"Grouped Rows", {"Column1", "Column2", "Column3"}, "Grouped Rows", JoinKind.LeftOuter),
    #"Expanded Grouped Rows" = Table.ExpandTableColumn(#"Merged Queries", "Grouped Rows", {"Count"}, {"Count"})
in
    #"Expanded Grouped Rows"

I would not recommend the final merge steps though. It is generally bad practice to leave dates in columns. They should be rows, both in Power Query and DAX.

But, it looks like this. This will work with any number of date columns.

How to use M code provided in a blank query:
1) In Power Query, select New Source, then Blank Query
2) On the Home ribbon, select "Advanced Editor" button
3) Remove everything you see, then paste the M code I've given you in that box.
4) Press Done
5) See this article if you need help using this M code in your model.