Release year: 2023


Buy this book (note: affiliate link)

Link to my handwritten notes


For a number of years, I have grown dissatisfied with the term “DevOps”. Don’t get me wrong; I am still incredibly thankful for the DevOps community that made me aware of a different way to see work and collaboration. In the past year, my understanding of DevOps shifted from a pipeline-centric approach to a learning-centric one. I fantasized about how we could use the DevOps way of thinking to streamline processes between teams that don’t have the names “Dev” and “Ops”, teams that might not even write code. I was daydreaming about using DevOps to solve problems happening between Marketing and HR teams.

However, not everyone agreed with me. I was recently told, “What you are talking about has value, but it is not DevOps. Please stop using the DevOps name in conjunction with your ideas, because you are hurting the movement and our efforts.” Ouch.

I believe the term “DevOps” has outstayed its welcome. I say this because I started noticing that depending on who you ask what the DevOps movement is working towards, you’ll get a different answer. This is a problem when you say your company “does DevOps”. What does that even mean? The problem here is an eternal one: naming things correctly is simply really, really hard.

I think authors Gene Kim and Steven J. Spear tackled this challenge head-on when writing this book. When Kim made DevOps mainstream with The Phoenix Project in 2013, he was actually showing us the tip of an immense iceberg. What he called DevOps at the time was actually a subset of a much broader perspective that could be applied to any type of work and problem-solving. It was time to give this movement a new name, something more general that could be used on a much bigger scale than only between Dev and Ops teams.

In their new philosophy gravitating around “Social Circuitry”, the authors give us new words to break free from the shackles of “Dev” and “Ops”. Here, we talk about the “winning zone”, the “danger zone”, coherence, coupling, the three layers of work, and the various “-ion” methods that help Layer 3 workers to wire their organizations with intent.

The three layers of work

As a student of Physics, I am seduced by this new way of thinking. It is as if Kim and Spear managed to write the unifying equations that tie all the different work improvement movements of the past in one simple, coherent toolbox. See in the figure below how they were able to generalize and simplify a multitude existing concepts through the lens of slowification, simplification and amplification:

How different practices slowify, simplify, or amplify

A person who was previously operating from a DevOps perspective now operates from “Layer 3”, the social circuitry, and has at their disposal the following tools:

  • Slowification makes solving problems easier to do (more time to think)
  • Simplification makes the problems themselves easier to solve (less complex)
    • Modularization partitions a large, complex system into coherent modules
    • Incrementalization breaks a few, complex experiments in many smaller, faster experiments
    • Linearization sequences tasks to flow successively
      • Standardization of tasks and exchanges at the delivery boundaries
      • Stabilization of problems to contain them and prevent spreading
      • Self-synchronization of the system without top-down monitoring
  • Amplification makes it obvious that there are problems that demand attention and whether or not they’ve been addressed

Overall, this is a brand new, deeply profound insight from two of the most acclaimed management gurus of our time. I am very excited to put their new thoughts to practice. Wiring an organization to win is probably the most important challenge in technology today, but thankfully we finally have a clear and coherent set of tools that allow us to get there.

Félix rating:

P.S. If I could make one suggestion to the authors, a simple change in the visual presentation of the different layers of work. It would be to picture Layer 1 work at the top and Layer 3 work at the bottom. I don’t like the visual presentation of having a “Layer 3 worker” live on top of a “Layer 1 worker”. Layer 3 supports Layer 1 & 2, and Layer 2 supports Layer 1. And Layer 1 are heroes.

⭐ Star quotes

  1. (p. 20) In all but the smallest endeavors, a leader’s primary contribution is not doing the work required to achieve the goal. Instead, they are responsible for everything required to enable that work to be done easily and well. This is achieved through the social circuitry by which people’s collaborative efforts are easily coordinated and integrated.
  2. (p. 41) The Nyquist-Shannon Sampling Theorem: A receiver (sensor) must sample at least twice the rate of the sender (the thing being monitored and controller) to accurately measure and control a system. Thus, if reports are generated and reviewed once a week, they can be used to control (manage) only situations that change no faster or more frequently than every 2 weeks.
  3. (p. 97) As a leader, are you humble and open-minded enough to expose your best ideas to aggressive testing? When those tests find flaws, will you be receptive enough to recognize that your best ideas have been refuted? Once those ideas are refuted, will you solicit the contributions of others to generate new ideas that can be tested?
  4. (p. 132) Multitasking is a source of cognitive load. It [invariably] degrades performance.
  5. (p. 251) Communication is not about speaking what we think. It’s about ensuring others hear what we mean.
  6. (p. 255) The best predictors of performance in tech environments:
    • to what extent important info can be shared
    • how messengers of bad news are trained to tell bad news
    • how responsibilities are shared across functional specialities
    • how bridging between teams is rewarded
    • how failure causes genuine inquiry