Something interesting happened the other day. Our company experienced a minor data incident that our data observability platform didn’t catch for the simple reason that the data was trapped in a silo. 

This got me thinking about other places data hides, not just from data observability platforms but from the scrutiny of the data team itself — and how these risks can impact our ability to successfully implement more modern or even experimental programs like data mesh. We sometimes refer to these as data silos, but really it’s the data equivalent of shadow IT.