Skip to content

Reading list

These are resources I’ve actually found useful — not because they’re famous, but because they change how you think and how you run the work.

This is the classic for product org design and what “good” looks like at scale. Read it if your team ships a lot but still feels like you’re not moving outcomes.

Continuous Discovery Habits — Teresa Torres

Section titled “Continuous Discovery Habits — Teresa Torres”

If you struggle to keep discovery lightweight and continuous, this is the playbook. It’s especially good for PMs who keep getting dragged into “big upfront research” that never finishes.

Short, sharp, and slightly painful — it teaches you how to avoid collecting polite lies. If your interviews feel “nice” but you learn nothing actionable, read this.

This is for teams stuck in output mode and calling it strategy. It gives you language (and courage) to shift conversations from “what are we building” to “what outcome are we driving.”

This is a different operating model: fixed time, variable scope, and a lot more discipline than most teams are comfortable with. Read it if your sprints keep turning into never-ending scope negotiation.

Lean Analytics — Alistair Croll & Benjamin Yoskovitz

Section titled “Lean Analytics — Alistair Croll & Benjamin Yoskovitz”

Helpful when you’re drowning in metrics and need to pick the ones that matter for your stage. It’s not perfect, but it’s a strong reset for teams measuring everything and learning nothing.

Positioning is not marketing fluff — it’s product clarity. This book is for PMs who keep hearing “it’s hard to explain what we do” and want a practical way to fix that.

Great for execution in complex environments where perfect plans fail on contact. It will make you better at creating direction without pretending you can control everything.

This is oddly useful for PMs because the job is basically attention management under stress. Read it if you want a pragmatic approach to building systems that survive busy weeks.

A solid, free resource for building AI features that respect human workflows instead of fighting them. If you’re working on AI products, it helps you avoid the “we added a model, now what?” trap.