How I Learn Product Management

Don't conflate PM with PM-the-job-title. The job title is a corporate construct. The skill is older than the title — it's what Steve Jobs, Tony Fadell, DHH, Andy Grove, Paul Graham do/did. Learn the skill, not the bureaucracy. The skill is: pick a real problem, design the smallest thing that solves it, ship, measure, iterate, kill what doesn't work, double down on what does. Everything else (OKRs, RICE, JIRA) is scaffolding.

See also Product Manager Books.md (comprehensive list, already curated) and Mental Models for Product Managers (12 mental models). This note is the method, not the resource dump.

Dependency graph

Product Management (the skill)
├── Picking the right problem  ← most leverage, least taught
│   ├── [On picking what to work on and which market to choose (startups)](/unpublished)
│   ├── [Strategies to find Niche Problems and Markets](/unpublished)
│   ├── [Competition is for losers](/unpublished) (Thiel)
│   ├── [blue ocean strategy](/unpublished)
│   ├── [Bizarre Behavior Signals Demand](/unpublished)
│   └── [Idea Maze of Private Data Sharing (genomics)](/unpublished)
├── Customer discovery
│   ├── Interview technique → The Mom Test (Fitzpatrick)
│   ├── Continuous Discovery Habits (Torres)
│   └── Jobs to be Done (Christensen)
├── Strategy & positioning
│   ├── [Peter Thiel](/unpublished) — Zero to One
│   ├── 7 Powers (Hamilton Helmer)
│   ├── Crossing the Chasm (Moore)
│   └── [Distribution as Moat and Competitive Advantage](/unpublished)
├── Solution design (the smallest thing that works)
│   ├── [wedge to expansion](/unpublished)
│   ├── Shape Up (Basecamp / [DHH](/unpublished) / Ryan Singer)
│   ├── [Focusing is about saying no - Steve Jobs](/unpublished)
│   └── [push complexity to the edges, keep the core primitive](/unpublished)
├── Prioritization & decisions under uncertainty
│   ├── [one-way vs two-way door decisions (of Jeff Bezos)](/unpublished)
│   ├── [prioritize ruthlessly](/unpublished)
│   ├── [Always rank the options decisions in front of you](/unpublished)
│   ├── [Kill the Risk Framework](/unpublished)
│   └── Thinking in Bets (Annie Duke)
├── Execution & shipping
│   ├── High Output Management (Andy Grove) — THE operator text
│   ├── Hard Thing About Hard Things (Horowitz)
│   ├── [The Primacy of Fast Feedback](/unpublished)
│   └── [feedback loops are extremely important](/unpublished)
├── Measurement
│   ├── [The Pirate Metrics (AARRR)](/unpublished)
│   ├── [The Arc PMF Framework](/unpublished)
│   ├── OKRs (Measure What Matters, Doerr) — read skeptically
│   └── North Star + guardrails (already in [Mental Models for Product Managers](/unpublished))
├── Behavioral / habit design
│   ├── [hook model (growth hacking)](/unpublished) / Hooked (Nir Eyal)
│   ├── Fogg's Behavior Model
│   └── retention-design lens
├── Taste & design
│   ├── [Taste](/unpublished)
│   ├── Design of Everyday Things (Norman)
│   ├── [UI Design System Concepts](/unpublished) / [UI Anti-pattern Concepts](/unpublished)
│   └── [Steve Jobs](/unpublished)
└── Communication (the multiplier)
    ├── Clear writing → see [How I Learn Writing](/unpublished) + [How I Learn Content Editing](/how-i-learn-content-editing)
    ├── [Clear writing free of fancy words is more composable](/unpublished)
    └── [Conciseness is a virtue in speech, writing, and code](/unpublished)

The single most under-weighted node is picking the right problem. Most PM books and courses are downstream of that choice. Spend disproportionate time at the top of the tree.

Books — practitioner × explainer Venn

Default to people who built hit products and can write. Reject professional-textbook authors and MBA-formulaic books.

Tier 1 — read first, partial chapters only (Don't read the whole book): - INSPIRED — Marty Cagan. The closest thing to a canonical PM text. Read Parts I–III, skim the rest. - BUILD — Tony Fadell. iPod / iPhone / Nest PM, deeply concrete, story-driven. Read cover to cover — exception to the don't-read-whole-book rule because it's compressed war stories, not theory. - High Output Management — Andy Grove. Operator's bible. Read Ch. 1–4 and the section on meetings. - Shape Up — Ryan Singer / DHH / Basecamp. Free online. Read it. The opposite of feature-factory PM. - Zero to One — Peter Thiel. Ch. 4 (Competition is for losers), Ch. 5 (Last mover advantage), Ch. 11 (Distribution).

Tier 2 — for specific nodes: - The Mom Test — Rob Fitzpatrick. ~100 pages. Read in one sitting. Single highest ROI book for customer interviews. - The Innovator's Dilemma — Clayton Christensen. Ch. 1–4 only. Skip the rest. - 7 Powers — Hamilton Helmer. Strategic moats. Short and dense. - Hooked — Nir Eyal. One sitting. Skim. Useful frame, not deep. - Design of Everyday Things — Don Norman. Ch. 1–2 are enough. - Continuous Discovery Habits — Teresa Torres. Cherry-pick interview rituals.

Skip / deprioritize: Lean Startup (Ries), Cracking the PM Interview, anything titled "X for Dummies" or framed for FAANG interview prep. These optimize for landing the job, not building the product.

The Tier 1 + Mom Test alone, partially read, beats reading every book on Product Manager Books.md.

Evergreen prerequisites

PM is a humanities-flavored discipline disguised as a corporate role. The evergreen prereqs are not "agile" or "SQL" — they're:

Writing clearly. PMs communicate constantly. See How I Learn Writing and How I Learn Content Editing. Untrained writers produce mush; trained writers produce decisions. Single biggest force-multiplier.
Reading humans. Behavioral psychology, motivation, why people lie in surveys. Mom Test + Hooked + Norman cover it.
Economic & strategic thinking. Incentives, Power of Incentives, market structure, Distribution as Moat and Competitive Advantage. Thiel + Helmer.
Decision-making under uncertainty. one-way vs two-way door decisions (of Jeff Bezos), Annie Duke, Kill the Risk Framework.
Engineering literacy. Enough to call BS on estimates and trade off complexity. Already covered.

See which knowledge is the most evergreen? sort them by fundamentalness — PM ranks near the top of the "humans + economics + writing" bundle, which is more evergreen than any framework you'll be taught.

Multi-perspective (HOW + WHY before WHAT)

In Baris's ranking order:

Historical / story — read Fadell's BUILD; Isaacson's Steve Jobs; The Hard Thing About Hard Things; Founders at Work (Livingston). How did the iPod, Stripe, Linear, Basecamp actually get made? Reject ahistorical framework-first books.
Computational / from-scratch — see "build from scratch" below. This is the most important step.
Algebraic / framework — RICE, ICE, AARRR, JTBD, OKRs, North Star. Learn the vocabulary; don't worship it.
Visual — opportunity solution trees (Torres), user story maps (Patton), funnels. Diagram by hand.
Verbal — Lenny Rachitsky's podcast, Shreyas Doshi, Garry Tan, David Sacks (All-In), Paul Graham essays. Background noise, not curriculum.

Build from scratch (the central move)

Apply What I cannot create, I do not understand—Richard Feynman to PM. The toy implementation = run a real PM cycle on a product you actually own.

Concretely, pick one project from the vault — SecureGenomics (Project), Vibecoded Directory Project, NS Search Project, or AutoResearch Project — and do the full cycle once, deliberately:

Discovery: 10 customer interviews using strict Mom Test rules. Record. Look for patterns.
Problem statement: one paragraph, no jargon. What job is the user hiring this for?
Wedge: what's the wedge to expansion — the smallest feature that proves the thesis?
PRD / RFC: write a 1-page spec. Use baris-agentic-coding-process if relevant.
North Star + 2 guardrails: pick the metric. Justify why.
Ship in ≤ 2 weeks. Cut scope ruthlessly. Take big bites. Anything worth doing is worth overdoing does not apply here — opposite is true for the wedge.
Measure for 2 weeks. Did the metric move? What did users actually do vs. say?
Kill or double down. Kill the Risk Framework. Most features should die.
Write up: atomic note in the vault, lessons learned.

One full cycle teaches more than five books. The books only become useful after the first cycle, because you now have hooks for the abstractions.

Anti-patterns

Reading 12 PM books before shipping anything. The 12-week plan in Product Manager Books.md is the wrong shape — invert it: ship in week 1, read in weeks 2–12 with questions.
Treating PM as feature-prioritization. PM is problem-prioritization. Features are downstream.
OKR theater. OKRs are useful tools; OKR rituals are bureaucratic cosplay.
Roadmap fetishism. Roadmaps are communication artifacts, not commitments.
Believing what users say over what they do. Mom Test rule #1.
Anemic Domain Models (design issue) applied to product: a PM who only manages tickets is anemic. Real PMs make trade-offs.
Learning PM by interviewing for PM roles. The interview is a different skill from the job.

Notes to write back into the vault

After the first cycle: - The Mom Test (book) by Rob Fitzpatrick.md - INSPIRED (book) by Marty Cagan.md - BUILD (book) by Tony Fadell.md - High Output Management by Andy Grove.md - Shape Up by Singer (Basecamp).md - 7 Powers by Hamilton Helmer.md - Thinking in Bets by Annie Duke.md - Jobs to be Done (framework).md - Continuous Discovery Habits by Torres.md - <Project name> PM Cycle 1 Retrospective.md — your own war story

Cross-link aggressively to existing notes: The Arc PMF Framework, The Pirate Metrics (AARRR), hook model (growth hacking), Mental Models for Product Managers, Distribution as Moat and Competitive Advantage, Idea Maze of Private Data Sharing (genomics), Strategies to find Niche Problems and Markets.

The one-line version

Pick a real problem you have. Interview ten people who have it. Ship a tiny thing in two weeks. Measure. Kill or double down. Repeat. Read Marty Cagan, Tony Fadell, Andy Grove, DHH alongside, never instead.