Design & Product Strategy
Mastery Guide

This guide is organized in eight parts, each building on the last. Parts 1-2 cover the strategic thinking and business fluency that sit underneath all product design at the principal level. Part 3 covers growth, engagement, and habit-forming design as a specific domain. Part 4 covers AI product strategy. Part 5 covers stakeholder management and organizational influence. Part 6 covers pressure-testing. Part 7 covers the principal-level transition. Part 8 is the preparation toolkit: frameworks, study plan, teardown method, interview prep, glossary, and reading list.

Design strategy is the discipline of making a small number of deliberate choices about what to design, for whom, why now, and at the cost of what else. It is not the design itself. It is the set of decisions that tell you which designs are worth making and which ones are not.

A working definition: design strategy is the connective tissue between where the business is trying to go, what the user is trying to accomplish, what the product can plausibly do, and what design should therefore prioritize. When any one of those four is missing, the result is not strategy. It is either wishful thinking, execution, or opinion.

Three common misunderstandings

Strategy is not a synonym for important. A large redesign is not automatically strategic. It is strategic only if it answers a real strategic question.

Strategy is not goals. "Increase retention by fifteen percent" is a goal. It says nothing about how, by what mechanism, at what cost. A strategy is a chosen path toward a goal. Two teams with the same retention goal can adopt opposite strategies (deepen the existing base versus broaden the top of the funnel) and both be reasonable, but only one can be right for a given company at a given moment.

Strategy is not a document. The strategy exists in the choices being made week to week: what gets built, what gets rejected, what gets deprioritized. If the choices are inconsistent, the strategy has not landed regardless of how well the deck reads.

The parts of a good design strategy

• A clear read of where the business is going and why now. Not a slogan, but the actual mechanism by which the company hopes to win.
• A specific point of view on who the user is, what they are trying to do, and where they are underserved.
• A short list of design bets that would move the business toward its aim by serving the user better than alternatives.
• An explicit set of trade-offs. What we will not do, and why.
• A statement of what has to be true for the strategy to work, in language specific enough to be proven wrong.
• A way to know whether the strategy is working, distinct from whether the designs are shipping.

Strategic thinking is not a personality trait. It is a set of habits practiced until they become automatic.

They ask why before what

When a request lands, the strategic designer traces it upward. Why is this being asked? What problem is this meant to solve? What would happen if we did nothing? What would happen if we did the opposite? Most requests are answers already, and the strategic move is to open the answer back into the question it came from.

They separate desirable, feasible, and viable

The classic IDEO frame. Desirability (does the user want it?), Feasibility (can we build it?), Viability (can the business support it?). Calling out which one is actually the constraint is often the highest-leverage move in the room.

They think in second and third order

First order: if we ship this, users will do X. Second order: because they do X, they will begin doing Y and stop doing Z. Third order: because Y is now common, the meaning of the product will shift and a new class of user or problem will appear. Most product decisions are made at first order. Most product regrets are second and third order. The habit: always ask "and then what?"

They read the room

Every organization has an official strategy and an operating strategy. The operating strategy is visible in what actually gets funded, escalated, and promoted. Design strategy that fits the official strategy but not the operating one usually loses.

They think in time horizons

Some decisions pay off in days (copy change), some in months (onboarding overhaul), some in years (platform bet). Strategic designers match the decision to its horizon and notice when leadership is asking for a short-horizon result from a long-horizon investment.

They surface assumptions

Every strategy rests on beliefs about the world that are almost never spoken. The strategic move is to write them down, mark which ones the strategy depends on most, and ask what evidence would confirm or disconfirm each. Assumptions turned into questions become experiments. Assumptions left as assumptions become surprises.

They notice patterns across domains

The way a subscription business retains users has structural similarities to how an enterprise SaaS keeps buyers renewing, which resembles how a nonprofit retains donors. Twenty years across enterprise commerce, regulated finance, and consumer growth is a pattern library, if you treat it as one.

Know how the money moves

Three questions matter for any company: How does it acquire users? How does it monetize them? How does it retain them? Reading a 10-K or S-1 for a company you are interviewing with is not optional at the principal level. It is the single fastest way to sound like a peer to the executives you will be talking to.

Understand unit economics

Unit economics is the small math of the business: CAC (cost to acquire one user), LTV (lifetime value), payback period, retention, conversion. A designer who can talk about how a proposed onboarding change would affect LTV/CAC will have a different conversation with the CFO than one who cannot.

Distinguish revenue, margin, growth, and moat

Revenue is what comes in. Margin is what is left after cost. Growth is how fast either is changing. Moat is what makes it defensible. Companies at different stages care about different ones.

Read the moat, not just the product

A moat is whatever makes it hard for a competitor to take the customer: network effects, switching costs, data advantages, brand, distribution, scale, or design taste that competitors cannot copy. Knowing whether design is the moat changes how much strategic weight design carries.

Watch what the CEO worries about

Every CEO has two or three things they worry about at any given moment, visible in all-hands, reviews, and board decks. A design strategy that speaks to those worries will be heard. One that is unrelated will be politely ignored.

The job, not the persona

The Jobs-to-be-Done question: what job is the user hiring this product to do, and what would they fire it for? The job is almost always emotional as well as functional. A user does not hire a fitness app to log workouts. They hire it to feel like the kind of person who works out.

Where the user is underserved

The strategic question is not "what do users want" but "what are users struggling with that the market is not addressing well?" Watch workarounds. What are users cobbling together across tools? Workarounds are the market saying "someone should design this."

The value exchange

Every product interaction is an exchange. The user gives something (time, attention, data, money). The product gives something back. At the strategic level, the question is whether the exchange feels fair and deepens over time. Products that die usually die because the exchange got worse for the user while getting better for the business.

The counterfactual

For every product, there is an alternative the user would use if this one vanished. A meditation app competes with Netflix and sleep, not other meditation apps. A budgeting app competes with a spreadsheet and denial. Naming the counterfactual honestly changes the strategy.

Segments, not averages

The average user is a fiction. Identify segments (power users, dabblers, lapsed, churners) and ask how each experiences the product differently. The strategic move is often to pick a segment to over-serve on purpose.

The alignment thesis

In a well-designed product, business and user goals are chained: earn attention by being useful, deliver value that creates a habit, let the habit produce engagement, monetize without breaking the habit, and let happy users bring in more users. Every link is both a user act and a business result. The design work: strengthen the weakest link.

The misalignment thesis

The chain breaks when the business asks for something the user does not want, or the user wants something the business cannot afford. The strategic designer's job: name the tension, quantify the cost of each side, and force a real choice.

Translating goals across layers

For any design decision, write a two-sentence chain: "This decision is meant to change [user behavior], which will produce [product metric], which will move [business outcome]." If the chain cannot be written, the decision is either not strategic or not thought through.

The metric trap and counter-metrics

Every metric is a proxy. When a metric becomes the target rather than the proxy, the strategy drifts toward optimizing the proxy while eroding the thing it was meant to measure (Goodhart's Law). For every primary metric, insist on a counter-metric that would rise if the primary were being gamed. Retention up but NPS down. Session count up but session depth down.

The product strategy stack

• Company mission: Why we exist. Changes rarely.
• Business strategy: How we plan to win. Changes every few years.
• Product strategy: What product bets deliver the business strategy. Changes yearly.
• Product roadmap: What we build in what order. Changes quarterly.
• Design strategy: How design serves the product strategy, for whom, at the cost of what.
• Design execution: The actual screens and flows. Changes weekly.

Most design work happens at the bottom two layers. Principal design work happens at layers three through five.

Product-market fit

Before PMF, the strategy is all about discovery: right user, right problem, right solution. After PMF, the strategy shifts to growth, retention, and monetization. How to recognize PMF: users complain when the product is down, growth is organic, retention curves flatten, the team can't keep up with demand. How to recognize its absence: growth requires constant spend, retention decays to near-zero, the team keeps pivoting the value proposition.

Product vision vs. product strategy

Vision is the future state of the world the product is trying to create. Strategy is the sequence of choices that gets there. Vision inspires. Strategy constrains. A common failure: compelling vision with no strategy ("we'll figure it out"), or strategy with no vision (efficient execution toward a destination nobody is excited about).

Positioning is a choice, not a description

April Dunford's framework from Obviously Awesome: competitive alternatives (what users would do without you), unique attributes (what you have that alternatives don't), value those enable (so what?), target customer (who cares most), and market category (what frame helps users understand you).

Design competitive analysis

Design competitive analysis differs from PM competitive analysis. Focus on:

• Experience gaps: Where do competitors' flows break down? What workarounds do their users employ?
• Design bets: What choices reveal their strategic priorities?
• Onboarding and activation: How quickly do competitors deliver value?
• Accessibility and inclusion: Are they leaving user segments underserved?

Prioritization inputs

The designer's role in prioritization

The most valuable design contribution is not scoring features. It is reframing the options. A PM presents a list of features to rank. A principal designer asks whether the right problems are on the list, and whether the proposed solutions are the only ways to address them.

Outcome-based roadmaps

The strongest product teams organize roadmaps around outcomes, not outputs. Instead of "Build feature X in Q3," the roadmap says "Improve day-7 retention by 5 points in Q3" and lets the team discover the right solution. This is the model from Escaping the Build Trap and Empowered.

North Star metrics

A North Star captures the core value delivered to users in a way that predicts business success. Good North Stars: weekly active creators, workouts completed, messages sent. The North Star should measure value delivered, not engagement extracted.

Leading vs. lagging indicators

Lagging indicators tell you what happened (revenue, churn). Leading indicators predict what will happen (activation rate, NPS, week-1 feature adoption). Designers should focus on leading indicators because those are the ones design can still influence.

Experimentation literacy

At the principal level, you should be able to frame a design change as a testable hypothesis, read an A/B test result, distinguish statistical significance from practical significance, and know when not to test: when the change is low-risk, when it is a values decision (accessibility, ethical design), or when testing would delay a time-sensitive opportunity.

These roles live inside consumer or prosumer products where success is measured not by whether a single task completes, but by whether the user comes back, engages more deeply over time, and builds a durable relationship with the product. This is the category behind Duolingo, Strava, Headspace, Notion, Spotify, and every learning, wellness, creator, and social product.

The central question is not "Can the user complete this task right now?" It is "Will the user still be using this product in thirty days, ninety days, one year?"

The mental model shift: funnels to loops

Enterprise commerce ends at a completed order. Growth products run loops: a user takes an action, the product responds with a reward, the user invests something back, and that investment increases the likelihood of the next action. The competitive pressure is also different: enterprise users open your product because their job requires it; consumer users open yours because it beat every other option in that moment, including sleep.

The core problems

• Acquisition and onboarding: Turning arrivals into activated users.
• Activation: The earliest moment where the product's value clicks. Highest leverage on retention.
• Habit formation: Reasons to come back without being annoying or manipulative.
• Motivation and progress: Helping users feel they are getting somewhere.
• Personalization: Making the product feel built for this one person.
• Community: Introducing other people so value grows with participation.
• Content discovery: Navigating a catalog larger than any user can browse.
• Activity tracking: Turning the user's data into a mirror they want to look at.
• Retention and resurrection: Bringing back fading users.
• Monetization: Converting engaged users to paying users without breaking trust.

The Hook Model (Nir Eyal)

Four stages: trigger (cues the user), action (smallest possible behavior), variable reward (payoff made compelling by unpredictability), investment (what the user puts in that improves the next loop).

Fogg Behavior Model (B = MAP)

Behavior happens when motivation, ability, and a prompt come together at the same moment. If a user does not act, one of the three is missing. The diagnostic tool for failing activation steps.

Self-Determination Theory

Humans are intrinsically motivated when three needs are met: autonomy (I choose this), competence (I am improving), relatedness (I am part of something). Products relying only on extrinsic rewards eventually break. This separates good gamification from bad.

AARRR (Pirate Metrics)

Acquisition, Activation, Retention, Referral, Revenue. Knowing which stage a project optimizes is the first question before opening Figma.

Growth Loops vs. Funnels

A funnel converts users once. A growth loop takes the output of one user and turns it into an input for the next. Referral loops, content loops, UGC loops, network loops.

Retention Curves

A healthy product shows a flattening retention curve. A "smiling" curve where retention rises later is the strongest signal of product-market fit.

Onboarding

The goal is not to teach the user everything. It is to get them to the activation moment, the smallest moment where value clicks. The canonical illustrations are historical, from the early-2010s growth era: Twitter's growth team (Josh Elman, circa 2011-2013) found that following roughly 30 accounts predicted retention; Facebook cited seven friends in ten days; Dropbox, uploading one file. Treat these as classic anchors for the activation-threshold idea, not as current metrics to copy. Principles: progressive disclosure, delayed sign-up walls, minimum viable onboarding. The best onboarding is invisible.

Habit formation

Habits form when a cue reliably produces a rewarded action, repeatedly, in a stable context. Streaks are the canonical mechanism but a blunt instrument. Design questions: what is the ritual, what is the smallest daily action, what is the reward that makes repetition worthwhile, and how do we handle a broken streak with grace?

Motivation and progress

Users continue when they can see progress. Progress bars, milestones, dashboards, weekly recaps, year-in-review summaries. Good progress design shows real progress on things the user cares about, not vanity metrics that celebrate the product's success rather than the user's.

Personalization

The design work is deciding what the system should learn, what signals to trust, how to explain choices to the user, and how to give the user control when the system is wrong. Explainability, control, feedback affordances, and recovery from bad recommendations are core problems.

Community and social mechanics

Once other users become part of the value, retention rises dramatically because leaving means leaving them. The tension: the same mechanics that drive engagement (public metrics, comparison) can produce harm at scale (anxiety, bullying). Modern community design is about norms and guardrails, not just affordances.

Notifications as a design surface

Every push, email, and in-app nudge is a decision about the user's attention. Good notification design: knowing when to send, what to say, what tone to use, when not to send, and how to give the user control. Consumer design lives and dies by this.

Activity tracking

Data is not insight. A chart of sleep is data. "You slept less on nights you drank alcohol" is insight. The best tracking products move up the stack from data to insight to recommendation to intervention. Each step is a distinct design problem.

What translates directly from enterprise

• Funnel optimization. Checkout conversion is the same discipline growth teams practice on onboarding and paywall flows.
• Data-informed design. The reasoning is the same; the tools change (Amplitude, Mixpanel, Heap).
• Design systems. Every consumer product needs one. Building one adopted across teams is directly transferable.
• Simplification of complexity. Consolidating fragmented workflows is exactly the mental muscle onboarding requires.
• Personalization thinking. Configuring, ranking, and defaulting in a complex catalog is adjacent to recommendation-system design.

What needs to be added

• Longitudinal thinking: Asking not "does this work in the session" but "does this build a relationship over months."
• Consumer sensibility: Tone, delight, motion, moments of surprise, emotional design.
• Behavioral psychology fluency: The vocabulary of habit loops, intrinsic motivation, variable rewards.
• Experimentation instincts: Framing design changes as hypotheses to test. Reading A/B results.

AI product strategy is not about building models. It is about deciding where probabilistic, adaptive systems create value for users that deterministic systems cannot, and designing the experience of living with a system that is sometimes wrong.

Where AI creates genuine design value

• Personalization at scale: Ranking, recommending, and adapting in ways manual rules cannot.
• Reducing repetitive effort: Auto-fill, smart defaults, draft generation, summarization.
• Surfacing hidden patterns: Anomaly detection, trend identification, predictive alerts.
• Natural language interaction: Search, chat, voice, document understanding.
• Creative assistance: Generating drafts and variations that the user refines.

Where AI does not belong

• Where the user needs deterministic, predictable behavior (financial transactions, medical dosing).
• Where the cost of being wrong is catastrophic and irreversible.
• Where the user has no way to verify the output and no recourse if it is wrong.
• Where the AI is a solution looking for a problem.

The confidence spectrum

• High confidence, low stakes: Act automatically. Let the user override. (spam filtering)
• High confidence, high stakes: Suggest, require confirmation. Present evidence. (fraud detection)
• Low confidence, low stakes: Suggest, with easy dismissal. (autocomplete)
• Low confidence, high stakes: Do not automate. Present data; let the human decide.

Progressive disclosure of AI

Start with suggestions (easily ignored), progress to drafts (easily edited), then to automations (easily overridden). Build trust before expanding scope.

Feedback loops

Every AI feature needs a feedback mechanism. Without it, the model cannot improve and the user has no agency. The best feedback is implicit: the user's natural behavior (clicking, ignoring, editing) is itself the signal.

Graceful degradation

AI features must degrade gracefully when the model fails. The user should always have a non-AI path to complete their task.

Users trust AI when the system is competent (usually right), transparent (they can see why), controllable (they can override), and aligned (it serves their interests). Trust erodes fastest when the system is opaque and wrong simultaneously.

User control patterns

• Preference tuning: Let users state what they want more or less of.
• History and audit: Let users see and edit what the system learned about them.
• Override and correction: Let users correct the system and have it learn.
• Off switch: Let users disable AI features entirely. The strongest trust signal.

Ethical guardrails

Test AI outputs across user segments for differential outcomes. Design review processes for AI-generated content. Build escalation paths for harmful recommendations. Be honest about what is AI-generated versus human-created.

Write before you present

A one-page memo will expose weaknesses that a twenty-slide deck can hide. Every design strategy should exist as a written document first.

Lead with the choice, not the analysis

Executives are choice-limited, not information-limited. Open with the recommendation and its consequences, then support with reasoning.

Make the trade-offs visible

Naming what will be given up is uncomfortable but produces buy-in that lasts. Executives are used to being sold to. Being level about costs is rare and memorable.

Frame choices, do not show designs

At the strategic level, showing pixels is a category error. Prepare the options being chosen between. Leading with designs invites feedback on the design. Leading with choices invites decisions.

Build coalition before the meeting

The decision meeting is where the decision is confirmed, not where it is made. The actual decision happens in one-on-ones and shared documents in the week before. Designers who wait until the meeting lose to designers who have already built agreement.

The stakeholder rotation

Before presenting, read your strategy from four seats: the CEO, the CFO, the head of engineering, and a user in the target segment. What does each see, and what would each push back on?

Align on outcomes, not artifacts

Alignment means everyone agrees on the outcome, the metric, the timeline, and the conditions for changing course. Agreeing on mockups without agreeing on outcomes produces the illusion of alignment.

Navigating disagreement

• Disagree and commit: State your position clearly, accept the decision, execute fully.
• Escalate with data: If the disagreement is consequential, bring evidence.
• Propose a test: If resolvable by evidence, propose an experiment with clear success criteria.

Unproductive moves: passive resistance, relitigating decided issues, going around the decision-maker.

What executives need from design leaders

• Clarity: Where design is strong, where it is weak, what the team needs.
• Candor: Honest assessment including when leadership's own decisions are creating problems.
• Business fluency: Connecting design decisions to business outcomes in executive language.
• Proactive risk identification: Flagging problems before crises, with proposed solutions.
• Low maintenance: Bringing solutions, not just problems. Executing without constant oversight.

The update format that works

Weekly or biweekly, one page or less: (1) What shipped and what we learned. (2) What is blocked and what I need. (3) What I am worried about. (4) What is coming next.

Working with product management

PM owns the "what" and "why"; design owns the "how" and "for whom." In practice, boundaries blur. The principal designer should be comfortable influencing the "what" and challenging the "why" while trusting PM to bring business context design cannot generate alone.

Working with engineering

The strongest partnerships are built on mutual respect for constraints. Designers who understand technical constraints make better decisions. Engineers who understand user constraints make better technical decisions.

Working with data science

Understand enough about models to ask: What data does this use? What populations is it biased toward? What happens when data is sparse? You don't need to build models. You need to design the experience of models being right and wrong.

Working with research

At the principal level, the relationship shifts from "commission a study" to "shape the research agenda." Identify the strategic questions research should prioritize. Ensure insights connect to product decisions, not just decks.

• Falsifiability test: State the strategy so it could be proven wrong. If it cannot fail, it cannot be right.
• Pre-mortem: Imagine the strategy clearly failed. What killed it? Surfaces risks group discussion misses.
• Steel man: Build the strongest possible argument against your strategy.
• Compared-to-what: Generate at least two credible alternatives. Ask why yours is better.
• Kill criteria: What would cause you to abandon the strategy entirely?
• If-we-succeed test: Imagine the strategy works. Is this world actually desirable?
• Evidence test: For each claim, name the strongest evidence for, strongest against, and evidence you wish you had.

• Clarity test: Look at the screen for three seconds. What is the user being asked to do?
• Failure mode audit: Empty state, error state, slow network, bad data, first-time user, screen reader user, non-native-language user.
• Dark pattern audit: Is the primary action easy and the escape hidden? Are we counting on cognitive load to produce compliance?
• Load test: Does the design work at 10x the users, content, and edge cases?
• Copy independence test: Does the copy make sense without the layout? Would a screen reader user understand?
• Friend test: Would you use this yourself? Would you send it to a friend and say "I made this"?

Strategic thinking traps

• Strategy as slogan. If it fits on a mug, it is not a strategy.
• Strategy as goals. The strategy is the how, not the what.
• Analysis paralysis. Using strategic thinking to avoid deciding.
• Being right in a way that does not matter. Analytically correct and politically dead.
• Believing your own story too soon. Every strategy is a hypothesis until evidence returns.
• Skipping the ethical read. Some strategies hit their numbers and produce harm. Naming this in advance is a strategic responsibility.
• Treating strategy as a document. It is the pattern of choices week to week.

Growth and engagement traps

• Gamification as the answer. The strongest signal of domain misunderstanding.
• Confusing engagement with value. A user who opens the app five times a day may be habituated, anxious, or trapped.
• Ignoring the ethics. Senior designers are expected to have a considered position.
• Skipping the quant. Cannot read a funnel or cohort chart = not taken seriously.
• Overclaiming. Name adjacent experience honestly.

• From "how" to "should we": Principals push back on the question, not just answer it.
• From delivery to direction: Judged on the direction the product moved because of your influence. Make direction legible: write it down, refer to it.
• From craft to taste: Making taste explicit in critique, written direction, design systems, and team standards. Silent taste is inaccessible. Legible taste scales.
• From peers to executives: Communication becomes compressed, choice-led, business-fluent, comfortable with ambiguity.
• From individual work to leverage: The highest-leverage act is often not designing anything. It is defining direction, choosing what not to do, and unblocking others.

• Read outside design. Business, finance, product, systems, psychology, history.
• Write regularly. Memos, position papers, strategic reads on companies. Writing exposes thin reasoning.
• Study failed strategies. BlackBerry, Kodak, Yahoo, Quibi. Failure stories are more honest.
• Practice the why ladder. Work upward through five whys to a strategic question worth answering.
• Do strategy exercises out loud. Pick a company you use and write their design strategy for the next six months.
• Find a strategy mentor. Strategic taste is transmitted through relationship more than reading.

Weeks 1-2: Foundations

• Read Hooked by Nir Eyal. Apply the loop to three products you use daily.
• Read first five chapters of Atomic Habits by James Clear.
• Read The Cold Start Problem chapters on network effects.
• Sign up for Lenny's Newsletter. Read the archive on activation and retention.
• Learn to read a retention curve, funnel, and cohort chart. Amplitude has free courses.

Weeks 3-4: Framework mastery and teardowns

• Teardown five products: Duolingo, Strava, Headspace, Notion, Reddit or Substack.
• For each: one decision to change, one to steal.
• Read Badass: Making Users Awesome by Kathy Sierra.

Weeks 5-6: Case study project

• Pick a real engagement problem. State a hypothesis. Design the solution end-to-end.
• Describe how you would test it and what could go wrong.

Ongoing

• One strategy book per month. One strategic memo per month.
• Constant product observation: what loop is running, what activation moment is targeted, what brings you back tomorrow.

A teardown answers these questions:

• What is the activation moment?
• What is the core loop? Name trigger, action, reward, investment.
• What is the retention mechanic?
• What signals is the product collecting and personalizing on?
• What is the tone and emotional design?
• What is broken or overreaching?
• What is the first thing you would test?

Two to three pages per teardown with screenshots. Three to four of these ready for interview discussion outperforms a bigger portfolio without domain fluency.

Questions to expect

• Walk me through a time you improved a conversion or engagement metric.
• How would you design onboarding for [product]?
• How do you decide what to test and what to just ship?
• How would you improve day-two retention for [product]?
• What is your point of view on gamification?
• How do you balance engagement metrics against user well-being?
• Data says users engage more with feature X, but research says it makes users anxious. What do you do?
• How would you evaluate AI-powered personalization vs. manual curation?
• Tell me about a time you pushed back on a product direction.
• How do you build alignment with stakeholders who disagree?

The domain-transition answer

"I have spent the last several years designing conversion and workflow experiences that lived and died on measurable outcomes at scale. The domain vocabulary is different in consumer engagement work, but the underlying craft of reducing friction, reading behavioral data, and shipping tested changes carries directly. I have been deliberately building fluency in the growth and engagement domain by [specifics], and I can talk in detail about [teardown or case study]."

Design exercises

Practice: state the metric first, the hypothesis second, sketch the solution third, describe the test fourth. This structure outperforms stronger visual work with weaker reasoning.

Core strategy shelf

• Good Strategy / Bad Strategy -- Richard Rumelt. Read first and twice.
• Playing to Win -- A.G. Lafley and Roger Martin.
• The Design of Business -- Roger Martin. Written for design leaders.
• Thinking in Systems -- Donella Meadows. Short, essential.
• Competing Against Luck -- Clayton Christensen. JTBD as rigorous method.

Growth, engagement, behavioral design

• Hooked and Indistractable -- Nir Eyal. Read together.
• Tiny Habits -- BJ Fogg. Atomic Habits -- James Clear.
• Badass: Making Users Awesome -- Kathy Sierra.
• The Cold Start Problem -- Andrew Chen.
• Drive -- Daniel Pink. Obviously Awesome -- April Dunford.

Product and design leadership

• Inspired and Empowered -- Marty Cagan.
• Continuous Discovery Habits -- Teresa Torres.
• Escaping the Build Trap -- Melissa Perri.
• The Making of a Manager -- Julie Zhuo.

Business and decision-making

• The Hard Thing About Hard Things -- Ben Horowitz.
• High Output Management -- Andy Grove.
• Thinking, Fast and Slow -- Daniel Kahneman.
• Nudge -- Thaler and Sunstein.
• The Innovator's Dilemma -- Clayton Christensen.

Newsletters and blogs

• Lenny's Newsletter. Growth.Design. Reforge Blog.
• Julie Zhuo, The Looking Glass. First Round Review.
• Andrew Chen's essays.