Dimensional Modeling: Banding

It's quite common to analyse a measure by categorizing it into different buckets or bands. For example to list the customers who are due for payment in the next week or fortnight, to decide which income group to target for a particular product campaign, which age group of students are most likely to buy a smartphone and so on. These bands are usually not supplied by source systems but derived from the underlying measure and they exist only for analytic purposes. While this banding could be achieved in reports or dashboards, there is significant value to model them as dimensions especially to promote self-service BI.