Data lineage is a record of where a piece of data comes from, which systems and transformations it has passed through, and everywhere it is ultimately used. It answers two questions that constantly arise within companies: why two figures about the same thing differ, and what will break if the source system or the method of calculation is changed. Without this overview, every change is made blind, and an error only shows up in a report that someone opens a month later. It does not need to be maintained with a complex tool – for smaller companies, a list of reports stating the source systems, the key calculations, and the responsible person is enough. Its value becomes apparent at the first major system change or the first dispute over the correctness of the figures.
See also: Data Warehouse, ETL and Data Flow, Data dictionary.
