The result of a query contains as many metadata columns as there are hierarchies on rows. For example, if we put two hierarchies on rows, the color and size of products, there will be two columns of metadata information, one for each hierarchy. In the first column, we will have all colors and in the second column we will have all sizes. Depending on the relationship between those hierarchies, we will get either a full cross join for unrelated hierarchies (different dimensions) or a reduced set of valid combinations (in case of the same dimension). In any case, there will be two columns.

Reports grow large very quickly. Once you put several hierarchies either on rows or on columns, suddenly you don't see data; all you see is metadata. That is, column and row headers. Sure, you can modify the layout of the pivot table, but still the report is not clear. Therefore, it's not uncommon that business users want to reduce a report's size, especially the width, by...