In addition to the advantages described above, however, key-figure dimensions also have certain negative aspects. Creating a measure for a single key figure somewhat complex because all key figures are in the same column which needs to be filtered in the DAX code. Settlement of key figures not registered in source data is also much more complicated, because calculations at the data-record level are not easy.
However, one more positive aspect of using a key-figure dimension is the tables' memory requirements. Although the additional dimension table also requires additional memory, this is compensated by the fact table's smaller size. In our example, however, this is not the case due to the small number of key figures and data records.
Increasing the number of key figures and entries per key figure would reverse the size ratio and relativize the additional memory requirement of the key-figure dimension. In an analysis of memory requirements, this is demonstrated by the, test table "FactContributionMargin Standarf extended" containing 20 additional columns and about 200 new rows: