An Elegant Puzzle by Will Larson, book cover
Leadership

An Elegant Puzzle summary

Systems of Engineering Management

The book in one sentence

Engineering management is a set of solvable puzzles, and you get better results by modeling your organization as a system and using repeatable methods than by reacting to each problem on its own.

Overview

Will Larson led engineering teams through fast, messy growth at companies like Uber and Stripe, and this is his field guide to the job. It covers the recurring puzzles of management, from how big to make teams to how to grow your own career. It reads like notes from someone who has been there.

Key idea 1: Treat the org as a system

Larson noticed that managers hit the same problems over and over and keep reaching for quick fixes that slide right off. His better approach comes from systems thinking, which he borrows from Donella Meadows' book Thinking in Systems. The idea is to model your organization as stocks and flows. A stock is an amount that builds up, like the number of engineers on a team or the size of a backlog. A flow is a rate that changes a stock, like how fast you hire or how fast people quit. Once you see the org this way, some things that felt confusing start to make sense. Take a short-staffed team. The obvious move is to hire fast. But new hires are a slow, costly flow. They take months to get productive, and while they ramp they pull time from the engineers who train them, so the team gets slower before it gets faster. A stock-and-flow picture shows that delay up front, so you stop expecting instant relief and plan for the dip instead. Attrition works the same way. If you only watch headcount, which is the stock, you can miss that a rising quit rate will drain the team no matter how fast you hire into it. You can even sketch these flows in a spreadsheet and play a change out over several months before you make it in real life, which is far cheaper than testing it on real people. The habit to build is this: when a problem keeps returning, stop patching the symptom. Draw the stocks involved and the flows that feed them, find the one flow you can actually move, and expect a delay before the results show up. That turns management from a string of reactions into something you can reason about ahead of time.

7 more key ideas in the app

  1. 2. Right-size your teams and managers
  2. 3. Know which of the four states your team is in
  3. 4. Migrations are how you manage tech debt
  4. 5. Lead change with model, document, share
  5. 6. Write strategies first, a vision later
  6. 7. Build every goal from four parts
  7. 8. Grow by choosing roles over rocket ships
Read the full Dib in the app

Did you know?

  • The book grew out of Larson's blog, Irrational Exuberance, at lethain.com, where he wrote about engineering management for years before Stripe Press pulled the ideas together and expanded them into the 2019 book.
  • Larson has led engineering at companies like Digg, Uber, Stripe, and Calm, and the examples come from scaling teams through hard, fast growth; he later wrote Staff Engineer and a guide for engineering executives.

Coming soon to the App Store and Google Play

DibsDabs for iPhone and Android is almost ready. Want to know the moment it's out? Send us a note and we'll tell you.

iPhone Coming soon Android Coming soon
Notify me at launch