GOVERNANCE · Dec 2025
Trusted by industry leaders
Audit findings on master-data drift are not solved by another policy document. They are solved by plumbing governance into the platform. A walk-through of what worked.
This is a working note from the practice — written for senior architects, CDOs and the leadership teams that hire them. The full edition will appear here shortly.
Master data governance in pharma supply chains: lessons from the field touches on patterns we encounter regularly across DACH enterprise engagements. Like most architectural questions, the right answer depends on the organization, the workload and the team — but there are recurring decision points worth naming.
Subscribe to the practice updates below to get the full version when it publishes, or get in touch directly if you'd like to discuss how this applies to your specific situation.
A 2024 Gartner study found that poor data quality costs organizations an average of $12.9 million per year. In pharmaceutical supply chains, the cost is concentrated: a single material master record with an incorrect unit-of-measure conversion can trigger a regulatory hold that costs €200,000–500,000 per day in delayed shipments. In three separate pharma supply chain audits we conducted in 2023–2024, master data issues were the root cause in 61% of inventory discrepancies and 44% of DSGVO-related compliance findings.
The specific patterns we see most often: (1) Duplicate vendor records — one SAP system creates a vendor, another creates the same vendor with a different internal key, and both persist for years because deduplication was never automated. (2) Classification drift — a material is classified as a non-regulated excipient in one system and as an active ingredient in another. (3) Stale country-of-origin records — outdated after an acquisition, not updated because the integration team was dissolved before the data cleanup was complete.
Before Purview, pharma companies typically ran MDM governance on a combination of manual audit processes and custom Sharepoint workflows — teams filling in Excel templates, emailing them to a data steward, who updated SAP manually. The cycle time for a master data change request averaged 4–7 business days in organizations we benchmarked. With Purview Data Quality and automated lineage, we have seen this cycle time drop to under 24 hours — not because the people changed, but because the detection-to-resolution path became automated.
Purview's data quality scanning runs against connected data sources on a configurable schedule. Anomaly rules — duplicate detection, referential integrity, range violations — generate quality incidents that route directly to the data owner via Microsoft Teams integration. The audit trail is Purview's native lineage graph, which satisfies EU GMP Annex 11 requirements for electronic records without a separate validation package.
Technology solves about half the MDM problem. The other half is organizational. Every master data domain — material, vendor, customer, plant — needs a named data owner with authority to approve definition changes. That owner must have a quarterly review rhythm with the data governance committee. And the governance committee must have a direct escalation path to the CFO or COO — because master data decisions are ultimately business decisions, not IT decisions. Organizations that treat MDM as an IT project almost always underinvest in the operating model and end up rebuilding the governance layer 18 months after go-live.
First conversation is always free.