97 Things Every SRE Should Know

Chapter 3. Building Self-Regulating Processes

Denise Yu

GitHub

In Camille Fournier’s excellent book, The Manager’s Path (O’Reilly, 2017), she advises readers to look for “self-regulating processes,” which caught my eye. My undergraduate degree is in economics, and I jump at any opportunity to apply economic thinking to practical problem-solving. Self-regulating processes are tiny cycles of checks and balances, and it’s cool to find them in human systems.

In my tech network, I often hear about process experiments succeeding or failing by the emotional or political bandwidth of the person who initiated the experiment. For example, when introducing pair-programming to a new group of engineers, it often takes a confident, charismatic person to coax reluctant teammates to start pairing for the first time.

In fact, they might not even call it pairing to begin with—they’ll say, “Hey, do you wanna come over here and have a look at this with me?” But when that person leaves a company, pairing might fall by the wayside, because it was something driven by the strength of a personality. These short-lived process innovations are valuable, but they don’t last; so in that context, we never learn how to adjust them, measure them, and scale them.

Self-regulating processes, on the other hand, don’t depend on strong personalities to persist. The way that they work is by aligning incentives (both ...