Finance says:
“That’s not how the process works.”
IT says:
“That’s exactly how the system is configured.”
The implementation partner says:
“That’s what was signed off.”
QA says:
“That’s what the test case says.”
Somewhere between those four answers is the process actually running in production.
Which version of the business are you testing?
An ERP test case can be technically correct.
And still be testing the wrong thing.
Because over time, something happens to every ERP.
- The documented process changes.
- A workaround becomes permanent.
- A configuration gets adjusted.
- A customisation is added.
- Someone builds an Excel step outside the system.
- A team starts doing something differently because that’s how they’ve always done it.
Eventually, you have four versions of the same process:
- The documented process.
- The configured process.
- The tested process.
- The process people actually follow.
Over time, those versions can drift further apart.
What I’d actually check
Not the test script.
The gap next to it.
I’d sit with the person who runs the transaction every day and watch them do it, not describe it.
I’d compare that against the configuration as it exists today, not as it was signed off at go-live.
And I’d ask:
When was the test case last updated?
Then:
When did the process last change?
Those two dates can tell you a lot about which version QA has actually been testing.
Most of the time, the answer isn’t that any of the four voices is wrong.
IT is right about the configuration.
QA is right about the test case.
The process owner is right about what they do.
They’re each accurately describing a different snapshot in time.
Why it matters
QA can execute every test successfully.
Automation can show green.
UAT can be signed off.
And production can still surprise everyone.
Not because nobody tested.
Because everyone tested a different version of the business.
So here’s the question:
Which version of your process would survive someone actually watching it get run?
