My 8 'Common Sense' Rules for Writing Better SQL

And One Annoying SQL Pet Peeve

Photo by Taylor on Unsplash

Proceed with caution — This is a label data does not come with.

It should.

Data has a way of keeping you humble —  just when you start to feel comfortable, something will come along and bring you straight back down to reality — normally hard and fast.

It should also come with an alarm bell, something along the lines of “do not operate heavy machinery.” Sadly, we live in this world, and in this world, it’s up to you to have your wits about you, and your shit together.

You’ve got to be careful not to let assumptions slam you into brick walls.

Simple formula I live by — Assumptions = bad. Planning and prep = good.

Always validate, always question, and always be prepared for something to go wrong — because data has a funny way of reminding you that certainty is just an illusion.

Take SQL, for example.

Once you’ve picked up the basics, you can easily fool yourself into thinking you “know” SQL (you don’t) — and it will make you learn the hard way the moment you plug your little query into a massive dataset — the game changes.

Maybe, with every man and his dog trying to plug in AI agents for SQL, things will get better? I doubt it — more murky, I’d wager. I’ve played with Gemini in BigQuery.