David Perell's Best Writing Advice: Lessons from 'How I Write'

David Perell has spent the last decade obsessing over one question: how do you write well on the internet? He turned that obsession into Write of Passage, the most popular online writing course in the world. He hosts the How I Write podcast, where he deconstructs the creative process with top writers. He built writingexamples.com to study great sentences and paragraphs the way athletes study game film.

Perell is not an academic writing theorist. He is a practitioner. He publishes constantly, tests ideas in public, and iterates on what works. His advice is specific, actionable, and field-tested by thousands of Write of Passage alumni.

Here are the key principles that run through his essays, courses, and podcast conversations.

1. Writing Is Thinking Made Public

This is the foundation of everything Perell teaches. Writing is not the act of recording thoughts you already have. Writing is the act of discovering what you think.

"If you can't write it clearly, you don't understand it." That is not a metaphor. It is literal. The act of putting words in sequence forces you to confront gaps in your reasoning. You think you understand a topic until you try to explain it in writing. Then the gaps reveal themselves. The vague parts become obvious. The assumptions you skipped over demand to be addressed.

This is why Perell argues that everyone should write, not just professional writers. Writing is a thinking tool. Engineers should write to clarify technical decisions. Founders should write to sharpen their strategy. Researchers should write to test their hypotheses before running experiments.

The implication for AI is direct. If writing is thinking, then having AI write for you means skipping the thinking. Bad writing is bad thinking, and outsourcing your writing means outsourcing your cognition. The value is in the struggle to find the right words. That struggle is the thinking itself.

2. Find Your Personal Monopoly

Perell defines a Personal Monopoly as the unique intersection of your skills, interests, and personality that nobody else can replicate. It is not your niche. It is deeper than that. Your niche is a topic. Your Personal Monopoly is the specific lens through which you see that topic.

A finance writer who also collects rare sneakers has a different perspective than a finance writer who plays competitive chess. Both cover the same topic. Neither sounds like the other.

The practical test: write about what you know that nobody around you knows. If you find yourself explaining the same concept to friends over and over, that is a signal. You have knowledge that is common to you and rare to others. That asymmetry is where your best writing lives.

Personal Monopolies cannot be replicated by AI. An AI can write a generic article about any topic. It cannot write from the intersection of your specific experiences, obsessions, and hard-won expertise. That intersection is your advantage. Lean into it.

3. Write From Conversation

Perell does not sit at a desk and try to think of ideas. He talks to people. Then he writes about what was interesting in those conversations.

He uses a framework called CRIBS to evaluate ideas during conversation. Watch for reactions that are: Confusing (people tilt their heads), Repeated (you keep coming back to the same point), Interesting (eyes light up), Boring (eyes glaze over), or Surprising (people say "wait, really?").

Confusing reactions mean the idea needs more work before you write about it. Repeated ideas are clearly important to you. Interesting reactions confirm the topic has an audience. Boring reactions are a kill signal. And surprising reactions are gold. Surprise means information asymmetry. It means you know something your audience does not. That is the foundation of a compelling piece.

The method is simple. Have conversations. Pay attention. Write about what provokes the strongest reactions. Your social life becomes your editorial calendar.

4. Find the Shiny Dime

Every great piece of writing has what Perell calls a Shiny Dime. It is one sentence that captures the core insight of the entire piece. One crystallized idea that a reader could repeat to someone else over dinner. If you cannot find the Shiny Dime, your piece is not ready.

Bad writing tries to say ten things. Good writing says one thing ten ways. The Shiny Dime is that one thing. It is the thesis statement stripped of all academic pretension. It is the tweet that could summarize your whole essay.

Perell recommends finding your Shiny Dime before you start drafting. Write the one sentence first. Then build everything around it. Every paragraph should either set up, support, or extend the Shiny Dime. If a paragraph does not connect to it, cut that paragraph.

This is one of those writing tips that actually works because it solves the most common failure mode: rambling. Most drafts fail not because the sentences are bad but because the piece lacks a center of gravity. The Shiny Dime gives you that center.

5. Build a Personal Information Architecture

Perell is obsessive about input systems. He argues that writer's block is usually an input problem, not an output problem. You are not stuck because you cannot write. You are stuck because you have nothing to write about.

The fix is what he calls a Personal Information Architecture. It includes your note-taking system, your highlight-capture workflow, your reading list, your conversation notes, and whatever tools you use to organize raw material. The architecture does not need to be complex. It needs to be consistent.

The key insight: you need a system that surfaces old ideas at the right time. A note you took six months ago might become relevant today. If that note is buried in a folder you never open, it is effectively lost. Perell advocates for systems that create collisions between ideas - random review sessions, linked notes, and regular synthesis of raw material into outlines.

This directly connects to the writing-is-thinking principle. Your notes are not just storage. They are raw material for thought. The act of reviewing and connecting notes generates new ideas. The architecture is not bureaucracy. It is a creativity engine.

6. Publish Consistently

The internet rewards consistency more than brilliance. Perell has said this repeatedly: write weekly, not when inspired. Inspiration is unreliable. A schedule is not.

Publishing on a schedule does three things. First, it builds an audience. People subscribe to writers they can count on. Second, it builds a body of work. One essay is a data point. Fifty essays are a portfolio. Third, it creates a feedback loop. You learn what resonates by watching what people share, what they respond to, and what falls flat.

Perell practices this himself. His Monday Musings newsletter has shipped every week for years. Not every issue is his best work. That is the point. Consistency beats perfection. You cannot improve at writing without publishing. And you cannot publish without accepting that some pieces will be better than others.

The practical advice: pick a cadence you can sustain and never break the chain. Weekly is ideal. Biweekly is fine. Monthly is the minimum. Less than monthly and you lose momentum, both for yourself and your readers.

7. Distribution Is as Important as Writing

Great writing that nobody reads does not matter. Perell is blunt about this. Most writers treat distribution as beneath them. They believe good work speaks for itself. It does not. Good work speaks for itself only after someone finds it.

Writing online means understanding how content spreads. It means writing titles that create curiosity. It means structuring pieces so they are easy to share. It means building an email list so you own your distribution instead of renting it from social media algorithms.

Perell distinguishes between writing for yourself and writing for an audience. Both are valuable. But if you want your writing to have impact, you need to think about the reader from the start. Not to pander. To communicate. A clear title, a strong opening paragraph, and a structure that rewards the reader's attention are not marketing tricks. They are basic courtesy.

This does not mean optimizing for clicks at the expense of substance. It means recognizing that substance without reach is a journal entry. Reach without substance is spam. The goal is both.

8. Study Great Writing Like an Athlete Studies Film

Athletes watch game tape obsessively. Musicians transcribe solos note by note. But writers rarely study other writers at that level of detail. Perell thinks this is a mistake.

He built writingexamples.com to change that. The site collects great sentences and paragraphs with annotations explaining why they work. It is a film room for writers. What makes this opening sentence hook you? Why does this transition land? How did the writer earn this emotional payoff?

The practice is simple. When you read something that stops you in your tracks, do not just admire it. Dissect it. Type it out word by word. Ask why each sentence follows the previous one. Ask what you would have written instead and why the original is better.

This kind of deliberate study compounds over time. You start noticing patterns. You internalize rhythms and structures. Your own writing improves not because you are copying anyone but because you have trained your ear to recognize what works.

Ambient Research

Perell estimates that 80% of his research happens before he sits down to write. Not in a dedicated research session. In daily life. Conversations, books, articles, podcasts, random observations. He captures these constantly and lets them accumulate.

He calls this ambient research. It is the practice of treating your entire life as input for your writing. The dedicated writing session is the last 20%. The other 80% happened in the weeks and months before, when you were not trying to write but were paying attention to the world. This is why Perell's Personal Information Architecture matters so much. Without a capture system, the ambient research evaporates.

Eighty Percent of Ideas Emerge After You Start

Perell cites Paul Graham on this point: 80% of your ideas about a topic emerge after you start writing about it. Not before. The act of writing generates ideas that thinking alone cannot produce.

This means waiting until you have the perfect idea is a trap. You will never have the perfect idea before you start. You will have 20% of it. The other 80% shows up during the draft. Start with what you have. The writing will surface what you are missing. This is the deepest argument for writing as thinking: the page reveals ideas that the mind alone cannot access.

Remixing Beats Originality

Perell does not believe in originality in the traditional sense. The best ideas are not invented from nothing. They are remixed from existing ideas, combined across fields that do not normally talk to each other.

A concept from biology applied to business strategy. A design principle from architecture applied to writing structure. A lesson from competitive gaming applied to career development. The value is in the connection, not in the individual components. Perell reads across disciplines specifically to find these connections. The wider your inputs, the more unusual your combinations, and the more original your output appears.

Escape the Never-Ending Now

Perell warns against what he calls the "Never-Ending Now." The internet rewards writing about what is trending today. The algorithm promotes recency. Feeds are chronological or engagement-sorted, both of which favor the new over the timeless.

The problem: trending content expires. A post about today's news is worthless next week. A post about a timeless principle is valuable for years. Perell prioritizes evergreen content. He writes about ideas that were true ten years ago and will be true ten years from now. This is harder than writing about trends because it requires deeper thinking. But it compounds. Every evergreen piece continues to attract readers long after publication. The Never-Ending Now is a trap that trades long-term value for short-term attention.

How AI Fits Into Perell's Framework

Perell's core thesis - writing is thinking - puts a hard constraint on how AI should be used. If writing is where the thinking happens, then AI should never replace the writing. The moment you hand the drafting to a machine, you lose the cognitive benefit that makes writing valuable in the first place.

But that does not mean AI has no role. The thinking happens in the draft. The polish happens in the edit. Those are different stages with different cognitive demands. Drafting requires you to struggle with ideas, find the right structure, and discover what you actually believe. Editing requires you to spot awkward phrasing, fix grammar, tighten sentences, and smooth transitions.

AI is excellent at the second task and dangerous at the first.

This is exactly the approach behind Athens. You do the thinking. You write the draft. Then AI helps you improve your writing through editing suggestions shown as inline diffs. You see every change the AI proposes. You accept or reject each one. The ideas stay yours. The prose gets sharper.

Perell's Personal Monopoly concept reinforces this. Your unique perspective cannot be generated by AI. It can only come from your experiences, your conversations, your specific intersection of interests. AI has no personal monopoly. It has the average of everyone else's writing. That average is competent and generic. Your writing should be neither.

Putting It Into Practice

Here is a workflow that combines Perell's principles with modern tools:

• Collect inputs constantly. Capture notes from conversations, books, and articles. Build your Personal Information Architecture. Use whatever tool sticks. The best system is the one you actually use.
• Test ideas in conversation. Apply the CRIBS framework. Notice which topics generate surprise and interest. Those are your next essays.
• Find the Shiny Dime. Before you draft, write the one sentence that captures your core insight. If you cannot write that sentence, you are not ready to draft.
• Draft without AI. Do the thinking yourself. Write ugly first drafts. Follow your curiosity. Let the structure emerge from the ideas, not from a template.
• Edit with AI. Once your draft captures your thinking, use AI to polish the prose. Review every suggested change. Accept what improves clarity. Reject what flattens your voice.
• Publish on schedule. Do not wait for perfection. Ship weekly. Build the body of work. Let the feedback loop teach you what resonates.
• Study what works. Read widely. Dissect great writing. Train your ear. The more you study excellent prose, the better your own drafts become, and the less editing they need.

The Core Lesson

Perell's work circles back to one idea: writing is a lever for thinking. Everything else - the Personal Monopoly, CRIBS, the Shiny Dime, consistent publishing - serves that central insight. Write to think. Think to write. The two cannot be separated.

In an age where AI can generate infinite text, the writers who thrive will be the ones who use writing as a thinking tool and use AI as an editing tool. The thinking is the value. The polish is the presentation. Get both right and you have something no algorithm can replicate: clear thought expressed in your voice.

For more on how to use AI as an editing partner without losing your voice, see our guide on how to improve your writing skills with AI. And for 30 concrete craft principles from Orwell, Klinkenborg, Zinsser, and Lamott, read 30 writing tips that actually work.