What I'd do differently
Master data problems are people problems with a data symptom — start with who owns the data, then build.
Insights
Notes on AI, MDG & delivery
Original writing — strategy, practice, and anonymized lessons. RSS feed
A lesson from the field
Most MDG projects fail in the quiet middle — design the change request process around how people actually work, then configure the tool.
Data governance is an ongoing business process
MDG succeeds when it is owned as a business process — Finance, Sales, Supply Chain — with data expertise in support, not the other way around.
Data migration vs governance
Migration gets your data in. Governance keeps it right — owners, approvals, validation, and replication for as long as the system lives.
MDG is not data migration
Data migration moves data once at go-live. MDG governs master data for as long as the system lives — owners, approvals, validation, and replication.