12 min · 11 chaptersNarrated in your voice
What it's about
An Elegant Puzzle collects Will Larson's working notes on running engineering organizations at fast-growing technology companies. Rather than offering leadership platitudes, Larson treats management as an engineering problem: teams, ratios, tools, and processes that either scale gracefully or break under growth. Drawing on his years at Uber and Stripe, he argues that most managerial failures are systems failures, and offers concrete heuristics for team sizing, measurement, hiring, and technical migrations that a working director or VP of engineering can apply directly.
Reading plan
Pick a finish date and Genius lays out the days — the plan shows today's target and keeps you honest.
Read together
Start a circle and share the code — everyone sees everyone's honest place in the book. Accountability, not leaderboards.
Who it's for
- Engineering managers and directors trying to move from instinct to explicit frameworks for running teams. - Founders and CTOs scaling an org past the point where informal coordination still works. - Individual contributors curious about what senior engineering leaders actually spend their time deciding.
What’s inside
- 1Management Failures Are Usually Design Failures1 min
Most breakdowns attributed to bad managers or lazy engineers are actually predictable outputs of a poorly designed system, so the fix is structural, not motivational.
- 2Team Size Is a Design Decision, Not an Accident1 min
The number of people on a team determines whether it can function at all, independent of how talented or motivated its members are.
- 3Where You Draw Team Boundaries Decides How Information Flows1 min
The way work is split into teams determines communication cost and ownership clarity more than any individual's effort, so org charts should be treated as an engineering artifact to be redesigned deliberately.
- 4Explicit Tools Replace Judgment as Organizations Grow1 min
Ad hoc manager discretion, which works fine in a small organization, becomes unfair and inconsistent at scale, so growing companies need explicit mechanisms like career ladders and calibration to standardize decisions at the cost of some flexibility.
- 5Most Engineering Metrics Measure the Wrong Thing1 min
Directly optimizing common engineering productivity metrics tends to make the organization worse, because the metrics are proxies that break down exactly when people start managing to them.
- 6A Manager's Calendar and Upward Relationships Are Infrastructure1 min
Controlling your own time and actively managing your relationship with your own manager are core management skills, not political extras, because a manager who does not shape the flow of time and information gets shaped by the organization's chaos instead.
- 7Retention and Sponsorship Matter More Than Hiring Volume1 min
Organizations chronically overinvest in hiring and underinvest in retention, and mentorship alone does little for underrepresented engineers without the distinct, harder-to-delegate work of sponsorship.
- 8Migrations Fail for Organizational Reasons, Not Technical Ones1 min
Large technical migrations succeed or fail based on how they are staged and incentivized across teams, not on the technical merit of the new system, so the real work is making the new path easier than the old one.
About Will Larson
Will Larson is an engineering executive who led infrastructure and platform teams at Digg, Uber, and Stripe before serving as CTO at Calm and later Carta. He writes the long-running blog Irrational Exuberance, and followed this book with 'Staff Engineer' (2021), both drawing on his own moves between individual-contributor and executive roles.
