#36 | Sunday reads for EMs

My favourite reads of the week to make your Sunday a little more inspiring.

Feb 22, 2026

Paid subscribers get 50 Notion Templates, The EM’s Field Guide, and access to the complete archive. Subscribe now.

One List to Rule Them All (James Stanier)

tl;dr: The word “priority” was singular for 500 years; we only started pluralizing it in the 1900s, and that linguistic shift gave us permission to avoid hard decisions. Maintaining a single stack-ranked list (no ties, no tiers) is a forcing function that externalizes debate and makes trade-offs visible - the discomfort of creating it is precisely the point, because it surfaces the conversations you’ve been avoiding.

Perfectly Designed for These Results (Mike Fisher)

tl;dr: “Organizations are perfectly designed to get the results they get” - and the corollary is that persistent problems aren’t people problems, they’re system problems. The piece pairs this with Conway’s Law to make the argument that your architecture literally mirrors your org structure, whether you designed it that way or not.

Three Bad Managers (Michael Lopp)

tl;dr: A vivid taxonomy of three archetypes. The Artist (visionary who doesn’t value humans), The Dictator (brilliant but bulldozes rooms), and The Knife (charismatically inscrutable) - all of whom were wildly successful leaders but terrible managers. Leadership is strategy and management is operations, and your boss almost certainly leans one direction; your job is to adapt to them, not change them.

Poor Deming Never Stood a Chance (Lorin Hochstein)

tl;dr: A discussion for why OKRs won and Deming lost. OKRs give bandwidth-constrained managers a tidy control system (set targets, monitor numbers), while Deming’s statistical process control requires a never-ending research program with no ceiling on what information might matter. OKRs aren’t better than understanding your system - they’re just easier, which is why they spread.

Do Less (Cate Hall)

tl;dr: Learning is digestion, not file transfer - cramming more in past a certain rate means nothing gets absorbed, and the real synthesis happens in downtime, not input. Optimizing your rest is still optimizing, and the internal “monitoring layer” needs to actually go offline for rest to do anything useful.

Living in the Inflection Point (Brittany Ellich)

tl;dr: The psychological reality of working in software right now. The cognitive overload of watching AI go from “science fiction” to “happening now” in 12 months. A useful observation: “glue work” (coordination, judgment, asking the right questions, knowing what to build) is becoming the primary skill now that AI handles much of the coding, which means the people historically overlooked for promotion are suddenly the most valuable people in the room.