In the management of structured data, as in the case of a Traditional Data Warehouse, the problem of reconciling data representing the same entity but coming from different sources must be faced.
Very often this activity is underestimated and managed in an unstructured way through the use of conditional rules (IF, CASE, etc.) inserted directly in the code on the Database and / or on the ETL tool: this situation often determines the loss of control over which , how many and where are the decodings applied in the data acquisition and / or transformation process.
In many cases, this translates into high costs for the operational and evolutionary management of the system.
As for me, I try to keep these 10 steps in mind when I am faced with the design of Data Mapping: even when it was not possible to apply the process in its interest, I have always had the opportunity to find benefits during the phases of ordinary, evolutionary and corrective maintenance of the system.
These advantages are particularly evident especially if they are compared with the system managed in this way with parts of the same system and / or mirror external systems, which use an unstructured management of the Mappings.