A heterogeneous dimension arises when different groups can be distinguished within a dimension, which have specific types of attributes per group. A product table can become quite large, there may be a phenomenon related to empty values. We don’t like the value NULL in a DWH, we prefer to use texts such as NONE or NA. But this can sometimes get out of hand if you have many different types of products, while there are actually attributes that are only relevant to 1 category. In this case we are talking about a heterogeneous dimension.