A version of this post first appeared in my newsletter. Subscribe to receive posts like this in your inbox every other Sunday.

I’ve written before about work and meta-work. In case you missed that post, whenever we’re working, we’re focused on one of two levels: the work itself and the meta-work that enables it. Developing effective work practices calls for balancing the two.

The work is whatever the task at hand calls for. For example, as I write this, my work consists of producing this issue of the newsletter. You can think of this as first-order work. Meta-work, or second-order work, consists of working on the systems that enable this first-order work. For example, I’ve set up processes and tools to help me write, edit, and publish this newsletter more effectively.

I’m always looking for ways to improve. For example, for a long time, I used Mailchimp’s web-based editor to build each issue of this newsletter. It was a time-consuming chore since Mailchimp’s editor isn’t primarily designed for authors. After much experimenting, I found a command-line tool that publishes Markdown files to Mailchimp. Adding this tool to my workflow has saved me dozens of hours.

Meta-work can significantly impact efficiency, but it can also become a drag. It took a long time for me to find the right tool, install it, learn how to use it, configure templates, etc. And it was somewhat scary: while building the new workflow, I wasn’t sure it’d pan out. It’s only retrospectively that you realize you’ve spent more time setting up “better” systems than the time they save you.

But at least this was a case where I sought to change the status quo — i.e., I could decide on the change’s timing. Sometimes, meta-work is thrust on us. For example, you could be starting a new job, which requires learning new communications systems. Or perhaps you need to replace a tool because the vendor is going out of business. Either must be done urgently.

Whatever the case, you should look out for how much time you’re investing in building and optimizing workflows. On the one hand, treating each new task as unique is wasteful; you should automate or ease repetitive tasks. On the other hand, it’s also possible to over-optimize systems and end up spending more time (and money) than they save.

The ideal is a balance between the two. This requires that you attune to how you’re doing first-order work. If it starts to feel like a repetitive chore, you should look for ways to set up systems and processes to help. Conversely, when you begin exploring such systems and processes, you should do so with the self-awareness to know when they’re becoming ends to themselves instead of supporting your first-order work. As with so many things, the key is mindfulness.