Been jammin' with @grantwinship about what separates great data teams. This is where we landed.

Great data teams realize they are in control of entity definitions.

What does this mean?

If I ask: "What does a record in the "customers" model represent?"

Most data teams: "We pull that data from our app"

Great data teams: "Someone who has purchased a product — they might come from our online store (our own app), or our retail locations (a POS system)"

Or: "What does a record in the "users" model represent?"

Most data teams: "That's from our users table"

Great data teams: "It's our best guess of one 'real life' person. They don't need to have an account — we included anonymous users in this table."

Most data teams: group reports / dashboards / analyses by data source — stripe, segment, zendesk

Great data teams: group reports / dashboards / analyses by business area — marketing, product, finance, support

How do we know this?

In my first data role, I let my source data define my data models — I made these mistakes myself!

We also see lots of companies we work with do the same thing. It feels like a subtle difference, but ends up being a huge paradigm shift.