This philosophy has a slightly more practical bent. Investing early in personal and project-based automation pays off compound interest dividends much later on. Personal productive automation that you invest in early on, and along the way as you work, are like compound interest. The earlier on you invest in defining the categories of things that can be automated, the faster you will find yourself running your project later on, and the fewer errors related to manual and out-of-order execution of steps you will find.

The easiest things to automate are oftentimes not the biggest. Rather, they are the small things that occur over and over. These are the things you should automate.

If you need an example, the zip archive of all of the notes that I make available to readers who purchase the eBook copy is automatically archived up and pushed to its hosted location inside a continuous deployment pipeline. Without this automation, I'd be stuck manually doing this step every time I update the knowledge base!