Start, sustain, and finish

How to finish a first draft fast — without losing the plot

August 16, 2026 · The StoryProp team

Ask how to write a book fast and most advice aims at your hands: type earlier, type longer, silence the inner editor. But the hands were never the problem. National Novel Writing Month's published target — 50,000 words between 1 and 30 November, an average of 1,667 words a day — is not many words, and nobody who misses it misses it for lack of typing speed. Watch a fast drafter and a slow one for a week and their keyboard speeds are nearly identical; what differs is the length of the pauses. The fast drafter has decided more — who wants what, what happens next, what the scene is for — and so spends the hour transcribing decisions instead of making them. That is the whole secret, and the rest of this piece is the practice of it: how to finish a first draft quickly without arriving at revision with a book that owes you a month of untangling.

Drafting speed is decision latency

Nobody drafts continuously, and the hour at the desk is mostly not typing at all. It is deciding whose scene this is, whether the argument happens in the kitchen or the car, whether Daniel would actually say the thing you need him to say. Typing is the cheap part. Deciding is nearly the entire cost. The most precisely documented drafter in English says the same thing from the other end. Anthony Trollope wrote with a watch on the desk and required 250 words of himself every quarter of an hour — three hours each morning from 5:30, the first half hour of it spent rereading yesterday — and recorded the results in his autobiography. That is a thousand words an hour, sustained, by a man who held a full-time job at the Post Office throughout; his self-imposed average was forty manuscript pages a week, 10,000 words, with recorded weeks running from 5,000 to 28,000. A thousand words an hour is not a typing feat. It is what the hour looks like when the deciding is already done.

So the fastest way to write a book is not to raise your words per minute but to shorten the gap between decisions. Two moves follow. Decide before you sit down: end each session with one line about what happens next, so tomorrow opens on a decision already made. And decide at the right size: “what is this chapter about” can eat a morning, while “what does Mara want in this scene, and who is in her way” can usually be answered in a minute — and the answer writes the next three paragraphs by itself.

This reframes the daily-word-count question too. If you want to write more words per day, don't audit your typing time; audit your open questions. A drafter who begins with the scene's want, obstacle, and exit already chosen will outrun a faster typist who begins with a blank. And when a stall stretches past a session into days, it is rarely a mood and usually a specific decision you have declined to make — a problem with its own toolkit.

The bad first draft has a limit

The permission slip everyone hands you is real: a first draft is allowed to be bad. “Mara stood in the kitchen and felt sad about the cup” is a placeholder, and revision exists to replace it with the image that does the work. Flat dialogue, thin description, a metaphor you already regret — none of it costs anything now, because prose repairs are local. You fix the sentence and nothing around it moves.

But the permission has a limit, and the limit is contradiction. If Daniel learns about the letter in chapter four and is astonished to discover it in chapter eleven, that is not a bad sentence; it is a broken fact, and every scene written on top of it inherits the break. Contradictions are not local. Repairing one means re-deciding which version is true, then rereading everything downstream to see what leaned on the wrong one. Ugly prose is a surface cost you pay once. A draft that contradicts itself is structural debt, and it compounds while you sleep.

Hence the working rule for speed: let the sentences be ugly, but never leave a load-bearing fact undecided. When you invent something at pace — a date, an injury, who knows what and since when — decide it completely and record it in the moment, while recording costs ten seconds. The prose can wait for revision. The facts cannot.

Never park at a chapter break

The clean stop feels virtuous: chapter finished, file closed, a small ceremony of completion. It is also the most expensive place to stop. Tomorrow the file opens on a blank page and a cold engine — new chapter, new scene, every decision unmade. The gap between chapters is where drafts stall for a week, because it is the point of maximum decision load and zero momentum. And they arrive often: by the working convention Reedsy publishes, an adult-fiction chapter runs 2,000 to 4,000 words, so at 1,667 words a day a chapter break lands roughly every one to two and a half days. That is a cold start several times a week, every week of the draft.

Stop in the middle instead — ideally the middle of a scene whose remainder you already know, mid-sentence if you can bear it. Then leave a downhill start: make the last act of every session one line about the first thing that happens next. “Daniel lies about where he found the cup; Mara watches his hands, not his face.” Tomorrow begins with transcription, and transcription warms into composition before resistance wakes up. Chapter breaks are for readers; session breaks are for writers. There is no reason they should coincide, and good reason they shouldn't.

Don't look back — except to leave a note

The classic sprint rule says don't reread. Yesterday's pages, revisited, become today's revisions, and mid-draft revising is how first drafts become eternal fourth chapters — a polished opening towing an absent book. Allow yourself the final paragraph for a running start, and no more.

The rule needs one exception, because sometimes the story genuinely changes behind you. Halfway through, you see that the flood should have happened in chapter two — it explains the damp in Mara's basement; it puts the letter in a water-stained box. Going back to write it breaks the sprint. Pretending you didn't notice poisons everything ahead. The right move is to log it and fix forward: one line in a running change list — “ch. 2: flood happens; basement damp from here on; the box is water-stained” — and from this sentence onward, write as if the change were already made.

Fixing forward costs a minute and buys two things. Everything you write from now on conforms to the corrected story, so your freshest material sits on the true version. And the change list becomes the first agenda of revision — a precise account of what the early chapters owe the later ones, instead of a month spent rereading to reconstruct it.

What writing a book in 30 days actually buys

The thirty-day challenge has a name and a number. National Novel Writing Month set the template — 50,000 words in the thirty days of November, an average of 1,667 words a day, checked by a word-count validator until 2019, after which winning became self-reported. So the sprint question is really two questions: what does it buy, and what does it defer. It buys a deadline that outruns self-doubt, public permission for rough prose, the discovery that you can produce at volume — which permanently resets your sense of what a writing day is — and, most underrated, a complete draft. A complete bad draft is a different object from a beautiful fragment. You can revise the first. You cannot revise the second.

The numbers also tell you where sprints break. In 2013, NaNoWriMo's best-documented year, 310,095 participants wrote 3.52 billion words between them and 42,221 validated a 50,000-word draft: a completion rate of 13.6 per cent, and an average across all entrants of about 11,350 words. The typical participant stopped at roughly a fifth of the goal — and nobody stops a fifth of the way in because their fingers gave out. The second thing to size honestly is the target itself. Using the Steve Laube Agency's rule of thumb, dividing word count by 300 for printed pages, 50,000 words is about 167 pages: a finished draft if you are writing middle grade or category romance — Harlequin's Desire and Presents lines both specify 50,000 words — and a healthy fraction of one otherwise. Current agent-stated targets, collected by the Manuscript Academy from agent Noelle Falcis Math of Transatlantic Literary Agency, put commercial fiction at 80,000 to 100,000 words and epic fantasy at 90,000 to 120,000 — which at 1,667 words a day is a 48- to 72-day sprint, not a 30-day one. The remedy is not to type faster. It is to set the finish line where your book's finish line actually is.

What a sprint defers is also real. Revision, obviously; that was the deal. But a careless sprint defers something worse: reconstruction. If you invented at speed and recorded nothing, the weeks after the sprint are spent reading your own book like a stranger's, working out what you decided and whether you kept to it. A sprint that saves four weeks of drafting and spends eight on untangling was not fast. It just moved the slowness somewhere with worse morale.

So take the loan — with the paperwork done: decided facts recorded, the change log running, a downhill start left each night. And if your life cannot host thirty consecutive days, most working lives can't; sustaining draft momentum on a crowded calendar is its own craft, and a slower drumbeat with the same bookkeeping finishes books just as surely. Trollope's 10,000 words a week were written before the Post Office opened.

Just enough continuity to keep what you saved

The continuity that protects a sprint is deliberately minimal — this is not the month to build a wiki. Keep two running documents, both append-only while you draft: the decided-facts list (names, ages, dates, injuries, who knows what and since when) and the fix-forward change log. Update each in the seconds after deciding, while it is a ten-second job; the same entry made three weeks later is an afternoon of rereading. Together they mean the finished draft arrives with a map — what is true, and what still needs making true.

This is exactly the bookkeeping a project-aware tool should carry, and StoryProp carries it. As novel-writing software built on a permanent project record, it turns what you agree to into records the agent reads before it writes — canon, characters, story shape, confirmed style rules — so “write it down while it's cheap” stops depending on your discipline at the tired end of a session. What you agree to gets written down. What you don't stays a suggestion.

A finished first draft is not a clean draft. It is a complete and decided one: every scene exists, every load-bearing fact has one version, and the known debts sit on a list instead of hiding in the text. Get there fast — momentum is a form of morale, and a deadline is a kind of kindness. What you are accelerating is the deciding. The typing was never the problem, and it was never the point.

A fast draft is only a bargain if the decisions you made at speed survive to revision. StoryProp writes down what you agree to as you go — canon, characters, confirmed style rules — reads the record before it drafts, and keeps every pass, so the month you saved isn't spent reconstructing what you meant.

← All posts