Civis Analytics

If you’ve heard the term "data warehouse," ETL is the moving crew that gets data there. Without it, the warehouse stays empty.

"We don’t need ETL, we just export CSVs." That is ETL. You’re just the ETL. And unlike a pipeline, you take vacation, forget steps, and occasionally paste into the wrong column.

Ever exported a CSV, cleaned it up, and uploaded it to another tool? That process has a name: ETL. 🧵

AI coding assistants are great until they hallucinate your table names. Civis Studio's MCP integration gives GitHub Copilot direct access to your actual data schemas. No more guessing. civisanalytics.com/resources/civi…

If your team is constantly copy-pasting between tools to build reports, a data warehouse is probably the thing you didn't know you needed.

"A data warehouse is just a big database" — nope. A database records things as they happen. A warehouse analyzes things after the fact. Different tools, different jobs.

You've heard "we need to get that into the data warehouse" and pictured a literal building full of hard drives. Let's clear that up. 🧵

Same thing we've seen with data tools for years: the orgs getting the most from tech aren't cutting headcount — they're investing in their people. AI widens the gap between teams that adopt and teams that don't. anthropic.com/economic-index

The heaviest AI users? Programmers, designers, writers, analysts — skilled workers building it into their workflow. The least usage? Both the lowest-wage AND highest-wage roles. AI isn't flattening the workforce. It's amplifying the middle.

Anthropic analyzed millions of AI conversations to see how people use it at work. 57% = augmentation (collaborating with AI) 43% = automation (delegating to AI) Only 4% of jobs use AI for 75%+ of tasks. The "AI is coming for your job" narrative doesn't match the data.

The 20 minutes isn't the query running. It's your colleague figuring out the right question to ask. The database is fast. Humans are the slow part.

"But can a query break something?" — a basic query just reads data. It's like searching a filing cabinet without moving any folders. You're fine.

Someone on your team said "I'll run a query" and disappeared into their laptop for 20 minutes. Here's what they were actually doing. 🧵