Workflow Wednesday: The Power of Small Working Examples

Art by @basilonmypizza: https://lnkd.in/eF8FkWzN - https://basilhefti.ch/

When ants search for food, they try paths, explore, and when they find something, they let the others know. A simple tactic. If you’ve ever had ants in your kitchen, you know how effective it is.

The same principle works remarkably well in data projects. Real understanding comes from working code. Slides make it easy to agree while completely missing each other. The moment you build something, everyone sees what’s possible, what’s needed, and what’s missing.

That’s the power of code sketches. They are our version of the ant’s first exploratory paths. A code sketch includes a minimal, idempotent setup for short create-use-teardown cycles, because names, schemas, and storage locations will change several times. It runs end-to-end, e.g. using Python and dbt, and validates its own results. Clean, minimalistic, focused.

From a distance, this code may look simple. On closer look, the clarity becomes visible. The sketch captures only what matters, and that clarity is what allows the design to evolve. Once a good path is found, the code is expanded to cover more cases and broader scope.

Do it like the ants: explore first, then refine. Work on concrete cases. Make sure your code runs from start to finish, and all parts fit together. Start with small, working code sketches. Iterate as your scope and complexity grow.

Where have code sketches helped you uncover misunderstandings early? Let’s discuss.

—

• Art by @basilonmypizza: https://lnkd.in/eF8FkWzN https://basilhefti.ch/

• Happy belated birthday, Julian Engel