Financial data governance is the set of rules that decides who owns each data area, what each term means, which system is authoritative, and how changes are made. Without it, every reporting obligation — statutory, tax, CSRD, Pillar Two — is answered by rebuilding data from scratch.
Governance has a bad reputation in finance because it is often delivered as a policy document nobody uses. The version that works is small, specific, and attached to real reporting obligations: a register of data areas, an owner per area, a definition per term, and a change process with a log.
For international groups the pressure comes from stacking obligations. The same underlying data now serves statutory accounts, transfer pricing, CbCR, Pillar Two and CSRD — each with its own definitions. Governance is what keeps those answers consistent with each other.
A lineage map answers 'where did this number come from?' in steps rather than in meetings. At minimum it records, per reported line item: source systems, transformation points, ownership at each step, and the controls applied.
It is also the artefact that makes system change safe. When a source system is replaced, the lineage map is the impact assessment.
Governance also covers who may see and change what. In practice that means role-based access aligned to the data areas, segregation between preparers and approvers, periodic access review, and personal data handled in line with the GDPR where HR and payroll data feeds reporting — which it now does under CSRD and Pillar Two.
The set of ownership, definition, control and lineage rules that make a reported figure reproducible: who owns each data area, how each figure is defined, which controls run, and where the evidence lives.
Controls confirm that a process ran. Governance ensures the data the process consumed is defined, owned and traceable, so the same number can be reproduced months later for an auditor, a tax authority or an assurance provider.