Comparison
Canonical datasetvsData mart
Canonical dataset
there is one table everyone agrees is the answer for orders, and the other four are explicitly labelled as derived or deprecated.
One agreed authoritative dataset for a given concept, with everything else derived from it. It is easy to say and organisationally difficult, because the competing copies usually exist for good local reasons and removing one costs somebody their workflow. The achievable version is not one table but one lineage root plus a clear statement of which derivations are sanctioned.
Full entry →Data mart
the finance team has its own set of tables, shaped for their questions, built from the shared ones.
A subset of the warehouse modelled for one team or one subject area. Built on conformed dimensions it is a useful presentation layer; built independently from source it becomes a second warehouse with its own definition of revenue, which is how organisations end up with four numbers for one metric. The distinction between those two outcomes is entirely whether the shared dimensions came first.
Full entry →