The line stops again.
You check ERP. Active supplier, current PO, everything looks fine. You ask the buyer. She says it’s covered. You call the plant. They’re waiting on a part that should have shipped last week.
You search the supplier name. Three results.
Same supplier, three different names. The legal entity in ERP. A shortened version in MES. A payables name that still includes an acquisition from 2019. And somewhere in Excel, because there’s always a version in Excel, a fourth name copied from a PDF that never made it into any system.
Now every question becomes a negotiation. Same supplier? Same account? Did we order from the right one? Did the ship date update hit the right record?
Nothing is technically broken. The work still gets done. But only because people are acting as the glue between mismatched IDs and duplicate records.
You’re paying plant managers to play detective with supplier names.
The Golden Record That Stopped Being Golden
The golden record stops being golden the moment updates happen elsewhere.
A supplier gets added in a rush. A site shortens the name to fit a field limit. AP keeps the legal entity. Purchasing uses the trade name. A ship date slips and someone updates it in Excel because that’s faster than opening a ticket.
None of this is wrong. It’s all normal.
It’s also how your golden record goes stale.
Once you have duplicates, the work shifts from operating to reconciling. Instead of asking “what’s true,” your team asks “which source should we trust?”
That’s when the golden record stops being golden.
When Every System Tells a Different Story
Duplicate suppliers are only part of it. The deeper problem is that each system keeps its own version of truth.
ERP holds the vendor ID and payment terms. MES tracks which suppliers are approved for which parts. AP cares about how the legal entity gets paid. Excel tracks whatever the last urgent change was.
None of these views are wrong. They’re just not aligned.
So the same supplier ends up with different IDs, names and statuses across systems. You can’t pay a supplier if planning thinks they’re on hold. You can’t plan if like 80% of companies, you don’t know where the accurate lead time sits.
Why Mapping Never Stays Finished
The usual answer is to map everything once.
You spend six months and six figures on consultants. They list the suppliers, line up the IDs, decide which system owns which fields, publish a master list. On paper, problem solved.
By the time they finish, your operations have already moved on.
ERP holds a single vendor ID. MES can hold different supplier codes by part or by line. AP needs the legal entity that gets paid, not the plant that ships. Each system was built for different users with different needs.
Mapping isn’t just matching columns once. It’s trying to pull all those local views into one picture that everyone can use.
But that picture is out of date the moment something changes:
- A new supplier gets added for a single part
- A plant gets acquired
- A site cleans up its own vendor list
If those changes don’t flow back to the master map, your golden record turns into a historical relic.
The Automation Gap
Here’s what most MDM approaches miss: it isn’t a data governance problem. It’s an automation problem. The mapping exists as a snapshot, not as a living sync.
A working golden record doesn’t just document the past. It needs to reflect the present.
When a new supplier hits MES, the record should catch it. When a name changes in AP, the record should flag it. When two systems disagree about the same supplier, the record should expose it before it stops a line.
But it can’t stop there. A golden record shouldn’t just read your systems. It should write back to them.
When MES calls them “John P” and ERP calls them “J Peters,” the record should recognize they’re the same supplier, not create a duplicate. When payment terms change in one system, the update should flow to the others. Automatically. With audit trails.
So the next person who needs that information gets the right answer without having to ask which system to trust.
That’s not how golden records work today. But it’s how they should.
What You’re Actually Paying For
When your golden record stays static, people fill the gap.
The cost of a static golden record isn’t just data quality. It’s plant managers debugging supplier names instead of managing a plant. It’s buyers reconciling records instead of negotiating better terms. It’s line leads waiting for someone to confirm which system they should believe.
That time adds up. Those delays compound. And eventually you stop seeing it as waste because it just feels like how things work.
It doesn’t have to.
Let the Record Keep Itself
A golden record that keeps itself current means the next time you search a supplier name, you get one result. Not three. Not four versions scattered across systems and spreadsheets.
One answer. The right one. Every time.
The line still stops sometimes. But not because nobody knows which supplier record to trust.
