Master data teams and solutions are so focused on fixing bad data, they’ve lost sight of the one thing that would actually solve it — preventing it from being created in the first place. And it’s failing everyone: the teams drowning in remediation work, the business users fighting bad data every day, and the organizations paying for a problem that keeps coming back.
Deduplication engines. Exception queues. Stewardship teams correcting records that entered the system wrong. Mass remediation projects that run every 12–18 months because the data degraded again.
For decades, car manufacturers ran what the industry called a “repair lot” — a section at the end of the assembly line where defective vehicles were parked, fixed, and reworked before they could be shipped. It was accepted as a cost of doing business. We’ve built the data equivalent of a repair lot and called it Master Data Management.
The costs are real — and most organizations only see half of them.
Direct costs are visible: MDM platforms, implementation projects, and stewardship teams whose primary job is fixing what should have been right the first time. Gartner estimates poor data quality costs the average organization $12.9 million per year.
Indirect costs are where the real damage happens. A bad material master doesn’t stay in MDM — it flows into procurement, manufacturing, finance, and logistics. A wrong unit of measure causes a mis-shipment. A missing classification blocks a product launch. A duplicate vendor generates a duplicate payment. These failures don’t show up in the MDM budget — they show up as operational rework, expediting costs, compliance failures, and lost revenue. Often five to ten times the direct cost.
If the cars keep arriving at the repair lot faster than you can fix them, the answer isn’t more mechanics — it’s fixing what’s wrong on the line.
Something is fundamentally wrong when fixing bad master data becomes a permanent line item.
Manufacturing figured this out decades ago.
Toyota’s Production System isn’t built around better defect correction. It’s built around defect prevention. Jidoka. Poka-yoke. The Andon cord. The philosophy is simple: stopping a defect at the source is always cheaper than correcting it downstream.
Toyota didn’t build a repair lot. They engineered quality into the process itself — and eliminated the need for one.
Master data has the opposite architecture. We put the repair lot at the end and called it governance.
The companies actually winning at master data aren’t better at cleaning it up. They’ve moved governance upstream — to the point of creation. Validated workflows. Business-owned approval gates. Duplicate detection before a record is saved, not after. Integration validation before bad data enters the system, not after it already has.
The goal isn’t a bigger repair lot. It’s a production line that doesn’t produce defects in the first place.
Instead of being a quality management system, MDM has become a quality control department.