Separate deciding what to say from saying it. A practical dump-group-order-expand workflow that makes drafting faster and structural fixes cheap.
· Productivity · 8 min read
Everyone who writes anything longer than an email knows the feeling. You sit down, you know roughly what the document needs to say, and the sentences still refuse to come. The usual diagnosis is a character flaw — not enough discipline, not enough coffee, "writer's block." The actual problem is more mechanical, and much easier to fix: you are trying to do two different jobs with one tool.
The fix is old, unglamorous, and works with remarkable consistency: outline first, write second. Not the ritualized outline from school, with Roman numerals and mandatory sub-points — a working outline, built in minutes, thrown away when it has done its job. This article lays out why drafting stalls in the first place, what a useful outline actually looks like, and a step-by-step workflow you can run on the next document you have to write.
Drafting from a blank page demands two kinds of decision simultaneously: what to say next — the structural question — and how to say it — the sentence question. They interfere with each other constantly. Every sentence you write quietly commits you to an order of ideas, and when that order turns out to be wrong, fixing it means tearing up sentences you already sweated over. So you hesitate before each sentence, because each one is secretly two decisions.
That interference is what a blank page actually is. "I don't know how to start" is almost never a sentence problem; it is a structure problem — you haven't decided what the first point is — being attacked with a sentence-level tool. No amount of staring produces the answer, because prose is the wrong instrument for the question.
Split the two jobs and both get dramatically easier. An outline is structural thinking with sentence-writing switched off: you decide what the points are and what order they go in, without spending a word on phrasing. Drafting then becomes sentence-writing with structure switched off: every paragraph already knows what it is supposed to say, so the only remaining job is to say it well. Neither half is hard on its own. It was the combination that was hard.
School outlines fail people because they are formalisms — numbered hierarchies you produce to prove you planned, usually written after the essay. A working outline is far simpler: one line per point, in the order you currently believe they should go. That's the whole format.
Two details make the difference between an outline that helps and one that decorates:
If you already keep notes in an outliner, this format will look familiar: it's just a bullet list where indentation means "supports the line above." That's no coincidence — outlining for writing and outline-based note-taking are the same motion pointed at different targets.
Write down every point the document might need to make — one line each, in whatever order they occur to you, with no filtering. Half-formed thoughts included. Ten minutes is usually plenty for anything short of a report.
The only rule at this stage is that no line gets rejected. Filtering while generating is the blank-page problem in miniature: two jobs at once. Generate now; judge later.
Drag related lines next to each other. Nest supporting details under the claims they support. This pass takes minutes, and it does quiet, useful work: duplicates reveal themselves, some "points" turn out to be evidence for other points, and the three or four real sections of the document emerge on their own instead of being invented up front.
Arrange the top-level groups into a sequence a reader could actually follow. All the classically painful questions — what to lead with, what the reader has to understand before point four makes sense, where the counter-argument goes — become drag-and-drop decisions here, instead of rewrite-three-paragraphs decisions later.
Read only the top-level lines, straight through, as if that were the piece. Does the argument hold? Is a step missing between two points? Is one section carrying nothing? This is the highest-leverage moment in the whole workflow: a structural flaw found here costs one drag to fix. The same flaw found in a finished draft costs an afternoon, and usually survives anyway because nobody wants to unpick two thousand polished words.
Now write — one bullet at a time, turning each into its sentence or paragraph. Two things make this phase feel almost suspiciously easy. First, you never face the question "what comes next?", because the outline already answered it; the blank page simply never appears. Second, you don't have to write in order. If the introduction feels hard, expand a middle section first — introductions are easier to write once the thing they introduce exists.
Promote the section lines to headings, smooth the transitions between expanded bullets, and delete any scaffolding that didn't earn a place in the prose. If you drafted inside an outliner, this step is mostly mechanical — in Indenta, for instance, "copy as document" converts an outline's structure into headings and prose in Markdown in one step, and the User Guide covers the details. But the workflow doesn't depend on any tool; a plain text file of one-line bullets works.
Structure stays cheap to change while structure is all there is. Moving a bullet takes two seconds and breaks nothing. Moving a paragraph means rewriting the transitions on both ends of the hole it left and both ends of where it landed.
Sunk cost never gets a vote. You will delete a bullet you spent eight seconds on without a second thought. You will fight to keep a paragraph you spent twenty minutes polishing, even after you've realized it doesn't belong — everyone does. Outline-first means the cut-or-keep decisions happen while everything is still cheap.
Momentum compounds. Drafting stalls when finishing one thought forces you to stop and decide what the next one is. With the outline in place, finishing a paragraph hands you the next prompt automatically, and the draft develops the rare quality of getting easier as it goes.
The honest counter-argument is that some writers discover what they think by writing, and an outline built too early can lock in a position they'd have written their way out of. The objection is real, but it argues for a looser grip, not for skipping the outline: do the discovering in the outline, where changing your mind costs a drag — and when the draft itself surfaces something better, follow the draft. Which leads to the failure modes.
Don't save the method for a novel. The next status update, proposal, or difficult email is the right test: spend ten or fifteen minutes on dump, group, order, and test, then expand. Even five bullets in front of a hard email changes how it goes.
Run it on two or three documents and the habit inverts on its own. Drafting straight into a blank page stops feeling like the brave, direct route and starts feeling like what it always was: doing two jobs at once, badly, when doing them one at a time was faster all along.
Try it in practice: Indenta is a free, offline-first outliner — nested notes, backlinks, tags, and peer-to-peer sync, with no account required. Start writing in your browser, or read the User Guide first.