To facilitate the analysis of historical trends, appropriate temporal data should be
included, at the desired granularity (e.g., daily, weekly, or monthly). For example,
suppose an operational data source stores the following fact type: Employee manages
Store. Over time, a store may have different managers. To retain history of these
changes, when store management facts are loaded into the data warehouse, the load
date may be inserted into the key, to populate the historical fact type: Employee on Date
managed Store. Basic aspects of temporal modeling were discussed in Section 10.3.