Watch Koe at his screen, recording for the first time, walking through how he actually makes a newsletter. He never opens a blank page and stares at it. By the time he reaches the "draft," the hard part is over — he's stringing together pieces that already exist, and he says the drafting "shouldn't take long, as the work is already done."1 The blank page, the thing every writer dreads, simply isn't in his pipeline. He removed it on purpose.
His line is the whole concept in one sentence: you don't write from a blank page, you write from ideas you've already stored.2 The four-stage process is the machinery that makes that true. A topic that performs, an idea dump, an outline where the real writing happens, and finally a draft that's mostly assembly.3 Most people experience writing as generating from nothing. Koe experiences it as arranging from a stockpile — and the stockpile is built continuously, on his phone, before any given piece begins.
Stripped of the app demo, the claim is structural: writing feels impossible because people try to do invention, organization, and articulation all at once, on a blank page, under time pressure. Koe splits those jobs into four stages and moves the heaviest one — organization — to the middle, the outline, where he says most of your writing actually happens.4 By the time you reach the draft, you're not inventing or even really organizing; you're stitching.
The four stages, in order. Topic — start from something that already performs on social, then make it your own, rather than guessing what might land.5 Idea dump — get every relevant idea out of your head and onto the page, knowing more will arrive later.6 Outline — the workhorse, where ideas get arranged into a persuasive, attention-grabbing order; this is the actual writing.7 Draft — fast assembly of the outline and the dumped ideas into prose.8 The genius isn't any one stage. It's that the dreaded blank-page moment has been distributed across four lighter steps, none of which requires creation from nothing.
Most people think the draft is the writing and the outline is a chore you do first. Koe inverts it. Picture a chef on a busy line: the cooking looks fast and impressive, but it's only fast because every ingredient was prepped, portioned, and arranged within reach during the quiet hours. The "draft" is the plating. The outline is the prep, and prep is most of the labor.9 If you've ever watched someone improvise a meal from a bare fridge, you know why blank-page writing is agony — there's nothing to reach for.
The deeper mechanism is the stockpile, and it runs underneath all four stages. Koe is blunt about the diagnosis: if you don't know what to write, it's because you haven't taken enough notes, haven't captured enough ideas, haven't written enough before.10 The four stages don't generate ideas; they organize ideas you already caught. Which is why capture matters so much — he singles out the mobile capture feature as the one thing made deliberately easy, because most people's best ideas arrive scattered, on their phone, throughout the week, nowhere near the desk where they'll eventually write.11 The pipeline runs on a fuel that has to be collected continuously, in motion, between sessions. The outline is where stockpiled fuel gets arranged; the draft is where it gets burned.
This page is the operational floor beneath a lot of the vault's more abstract claims about ideas and notes. The idea museum / swipe file argues for hoarding raw material; this page shows where in the writing process that hoard gets spent — the idea dump and the outline pull directly from it. Without a process to discharge the swipe file, a swipe file is just a graveyard. This is the discharge mechanism.
It also gives the persuasion and framework pages a place to live. Koe's frameworks (PP, PASO, PASTOR) are what you apply inside the outline stage — they're the ordering logic for arranging dumped ideas into a persuasive sequence. So this page is the container; those are the shapes you arrange inside it. And it hands the AI-collaboration corner of the vault a precise hook: the whole pipeline is built on stored, retrievable ideas, which is exactly where the AI second-brain retrieval shift enters.
The clearest case is Koe's own outline template, which he scrolls through on screen — because it shows what "most of the writing happens here" concretely means.12 His outline isn't a skeleton of headings. It's a structured workspace of "elements": a brain dump, a list of potential hooks researched from his past videos and tweets, a connections panel for pulling stored ideas, a research element where he embeds the specific tweets he'll reference, and then a block where he fills out — for the chosen topic — the problems, the goals (the opposite of the problems), personal experiences, examples, and a process.13
Notice what that last block is. Problems, then amplification, then solution — he says so directly: "you can think of this as problem and amplify and then solution."14 The outline stage is where a copywriting framework gets poured into the raw idea dump. He then distills all of it into a "key points" element that he writes from.15 By the time the draft exists, the topic is chosen, the hook is researched, the supporting tweets are embedded, the pains and solutions are mapped, and the key points are extracted. The draft is reading his own outline "word for word" and leading into what comes next.16 His best-performing videos — the one-person business, "how to productize yourself" — were, he says, just the concept and process elements in his outlines.17 The case study proves the thesis: the outline isn't preparation for the writing. It is the writing.
It's early, still dark, and you're not at your desk — you're on the edge of the bed reaching for your phone, because an idea just surfaced the way ideas do, sideways and unannounced. You don't trust it to survive the commute. You open capture and thumb out one line, tag it tweet idea, and put the phone down. It joins forty others from the week, each one caught in motion, none of them at the desk.
Tuesday, writing morning. Now you sit down — but not to a blank page. You pick a topic you saw perform on a feed you trust and decide to make it yours. You open the outline. The brain dump goes first, fast and ugly. Then you pull the week's captures: you tap the tweet idea tag and there they are, the forty stray thoughts, filtered and waiting, and three of them slot straight into this piece. You fill the problem-and-goal block, drop in the two tweets you'll cite, distill the key points. The morning's real work happens right here, in this arranging.
Only then do you open the draft. And the strange thing — the thing Koe promised — is that it goes fast, almost dull, because there's nothing left to invent. You're stringing. The blank page never showed up, because you killed it at 6:40 a.m. with a tagged one-liner on your phone.
Single popular-source transcript, [PARAPHRASED] [POPULAR SOURCE] throughout. Most of the evidence is Koe narrating his own screen, which is unusually concrete for a creator transcript — you're watching the actual template, not a summary of it. That raises the reliability of the description of his process. It does not, by itself, establish that the process generalizes.
🚩 MOTIVATED REASONING — the heavy kind. This entire walkthrough is a product demo for Kortex, his note-taking app. The four-stage process is presented inseparably from the tool that's claimed to enable it (capture, tags, connections, elements, panes). Koe does say, once, that you can technically recreate this in any note-taking software.18 Keep that caveat load-bearing: the pipeline (topic → dump → outline → draft, fueled by continuous capture) is the transferable claim; the app is the sales pitch wrapped around it. The vault files the pipeline, not the product.
Open questions. Koe's "if you don't know what to write, you haven't captured enough" is a clean diagnosis, but it risks blaming the writer for what might be a retrieval failure rather than a capture failure — a buried stockpile and an empty one produce the same symptom. He also leaves unexamined how much the "topic that already performs" step quietly biases the whole pipeline toward imitation of what's already trending. [PLAUSIBLE — needs corroboration] on the universality.
Koe and the idea-museum / swipe-file page agree almost completely on the front end and diverge interestingly on the back. Both insist the writing depends on a continuously-built reservoir of captured material — neither believes in generating from nothing. Where Koe adds something is the discharge path: the swipe-file page is strong on collecting and weak on spending, and Koe's four stages are precisely the spending mechanism, with the outline as the conversion chamber. Read together they form a full loop the swipe-file page only half-draws: capture continuously, then dump → outline → draft to turn the hoard into prose.
There's a sharper tension with the creative-practice pages on ideas-in-motion — the cluster that says ideas are caught while walking, in the body, away from the desk. Koe's pipeline is desk-bound and systematic; it can read as the opposite of "trust ideas will come back, don't capture everything." But the contradiction is softer than it looks. Koe's capture step lives exactly where the walking pages live — on the phone, in motion, between sessions — and his pipeline step is what happens after, at the desk. The unresolved edge is philosophical: the walking pages would say capturing everything is itself a kind of grasping that kills the wandering mind, while Koe would say an uncaptured idea is a lost one. Both can't be fully right, and the page keeps the friction rather than smoothing it. The likely truth is that which ideas you trust to return, versus which you capture, is the real skill — and neither side fully specifies it.
Plain version: this page is about where ideas get stored and arranged before they become writing — and that storage-and-retrieval question is the exact place where AI tools, second-brain systems, and the craft of editing all plug into the writing process.
The first handshake is with the AI second-brain retrieval shift. Koe's whole pipeline runs on stored, retrievable ideas — tags, connections, captures you pull at outline time. That's a retrieval system built by hand. The AI-retrieval page argues the bottleneck in second-brain work is shifting from storage to retrieval — that the value is no longer in hoarding notes but in surfacing the right one at the right moment. Set the two together and a sharp implication appears: the stage of Koe's pipeline most exposed to AI is not the drafting (which everyone assumes) but the outline — specifically the moment he taps a tag and pulls his stored ideas into the workspace. If retrieval gets dramatically better, the outline stage gets dramatically cheaper, and Koe's "most of the writing happens in the outline" might quietly become "most of the retrieval happens in the outline." The human work would migrate again — from arranging what you can find to judging what gets surfaced. That migration is something neither page sees alone: Koe describes a manual retrieval craft; the AI page describes the force about to automate it; the overlap shows you which stage is load-bearing and therefore which stage is most at risk.
The second handshake is with the beta / editor review skill. Koe's four stages end at "draft, then post it so you can get data and improve."19 There's a gap there — between draft and post — that the editor-review page fills: the discipline of reviewing and tightening a draft before it ships. Put them side by side and you get a five-stage picture Koe's transcript only implies: topic, dump, outline, draft, and review. The interesting tension is that Koe's pipeline is optimized for speed (the draft "shouldn't take long") and the editor-review skill is optimized for quality (slow down, audit, cut). Both are right at different scales — a daily social post can skip the editor stage and let the audience be the editor (Koe's "post it and get data"), while a high-stakes landing page can't. The insight in the overlap: the four-stage process tells you how to generate a draft cheaply; it's silent on when generation is enough and when judgment has to be added, and the editor-review page is exactly that missing decision. Speed-to-draft and review-before-ship aren't competitors; they're the two halves of a complete pipeline, and Koe ships you only the first half.
The Sharpest Implication. The blank page isn't a creativity problem — it's a logistics problem you can engineer away, and the engineering happens days before you sit down, on your phone, in motion. If that's right, then "I can't write" is almost never about talent or inspiration; it's a symptom of an empty or unsearchable stockpile. The lever isn't writing harder at the desk. It's capturing continuously away from it, and building an outline stage strong enough that the draft is just plating.
Generative Questions.