Standardization looks tidy until exceptions pile up

I keep noticing that platform standards feel clean right up until a team has one weird workflow and starts bending them. Then the shared system stays “simple” on paper, while everyone quietly builds little escape hatches around it.

Has anyone found a good line between team ownership and platform control that doesn’t turn into a mess of exceptions?

“Escape hatches” usually means the standard didn’t leave a real lane for the weird-but-legit workflows, not that the team is being difficult.

When the platform never clearly says “exceptions live here, and here’s how we review them,” people just build their own side doors and call it “temporary”… forever.