(Photo by Jose Fontano on Unsplash)

Underappreciated thought in the data world: Data silos are … usually a good thing!

By that, I mean:

It pays to be thoughtful and intentional about sending data beyond the team/department/area where it was collected.

Simply dumping it all into a big data lake – a free-for-all where anyone can grab anything, without asking and without stating a clear purpose – opens the door to a variety of issues down the road.

This is when you'll remind me that mixing datasets is how you surface interesting analyses and insights. This is true!

It's also true that some of those "interesting" finds raise technical, ethical, and possibly regulatory concerns.

When you're careful about letting data travel, those problems are harder to come by.