#40 | Sunday reads for EMs

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

Stephane Moreau

Apr 12, 2026

👋 Hey, it’s Stephane. Every Sunday I share with you my favourite reads of the week.

To accelerate your growth see:

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

Why Your Engineering Team Is Slow (Ally Piechowski)

tl;dr: Introduces “codebase drag” and gives you a 5-signal scoring rubric (apology estimates, deploy fear, “don’t touch that” files, the coverage lie, time to first commit) to diagnose it. Worth reading if you’ve ever had to justify a refactor sprint to your product manager - it gives you a vocabulary and a number to put in front of them instead of vague “tech debt” hand-waving.

Encoding Team Standards (Rahul Garg)

tl;dr: Argues that AI coding instructions should be treated like CI configs or lint rules (versioned, reviewed, repo-resident artifacts) not Slack tips or tribal knowledge in seniors’ heads. The core insight is that AI-assisted development turns inconsistency between juniors and seniors into a systems problem you can actually solve: extract what the senior instinctively prompts for (architectural constraints, threat model, refactoring philosophy) and encode it as executable instructions with role/context/categorized-standards/output-format anatomy.

The Alarm That Went Silent (Mike Fisher)

tl;dr: Teams optimize for the metrics they can see, while risk builds in areas that are harder to measure: team burnout, frustrated customers who haven’t churned yet, fragile systems, weakening culture, or slower decisions caused by too much process. Even more importantly, teams rarely stop to ask a harder question: how would we know if our metrics stopped telling the truth?

Say the Thing You Want (Matheus Lima)

tl;dr: Why engineers (and engineering managers) sit on what they want in 1:1s and how that tanks careers - managers have seven other reports and incomplete information, so an unspoken desire has zero surface area for anyone to help with. People confuse wanting something with working toward it, and saying it out loud to another person is what actually gives the goal weight and changes your behavior.

On the Socially Acceptable Use of AI in Business (Dave Kellogg)

tl;dr: Use AI however you want, but you are responsible for the result. Rejecting an idea just because AI helped create it misses the point. What matters is whether the idea is good. The real problem is presenting AI output as your own thinking when you haven’t read it carefully, challenged it, and corrected it. AI can help you work faster, but you still need to understand the work. If you let AI do the thinking for you, the mental skill you’re supposed to build will slowly weaken.