When you add a column to a pivot table or to a slicer, by default, the data is sorted in the ascending order either by digit or by character. This is what you can see on the Month attribute in the preceding screenshot. In a lot of cases, this is not what you want to do; instead, you want to sort it based on another attribute, in the month example, you typically want it sorted by the month number instead.
When working with multidimensional models, you can solve this by the NameColumn, KeyColumn, and OrderBy attributes of the attribute. However, in tabular models, there is no concept of name and key columns; so how do you solve this?
In the first version of PowerPivot, there was no other option than to create an attribute that looked like 01_January to sort January as the first month. However, as one of the new features in SQL Server 2012 the tabular model has an option to sort a column by another column. This gives the developer the ability to mark a column...