You sat down at nine with two free hours and a chapter to write. At eleven you have four hundred words, three of which you like, and the uneasy feeling that the time went somewhere you cannot account for.
It did, and it was not typing. A competent typist produces four hundred words in about eight minutes. Writing slowly is almost never a speed problem — it is a decision problem, and the decisions are being made at the worst possible moment, one sentence at a time.
Account for the two hours honestly
If you log a slow session minute by minute for a week — most people only need to do this once — the same four columns appear, and typing is the smallest of them.
Deciding what happens
the largest columnStaring at the screen working out whether they argue in the car or at the house. This is planning, and it is being done in the most expensive place possible.
Rereading and polishing what exists
the sneakiestScrolling up to fix a word in yesterday's paragraph, then the one above it. It feels like work, produces nothing new, and will happen again tomorrow on the same paragraphs.
Looking things up
the most disguisedWhat year did the bridge open, what does a paramedic actually say on arrival, what is the Arabic for this. Forty minutes later you know a great deal and have written nothing.
Actually producing new sentences
often under 20%The only column that moves the book. Every technique below exists to make it bigger at the expense of the other three.
Four changes, in order of effect
One
Never open a session without knowing what happens in it
The single biggest change available to most writers, and it costs three minutes. Before you stop today, write one line at the top of tomorrow's page: she tells him about the money, he pretends he already knew, the meal ends badly. Not an outline of the book — a note about the next two hours.
Tomorrow you open the file already knowing the destination, and the sentences arrive at a completely different rate. Hemingway's version of this was stopping mid-scene while he still knew what came next, so the next morning began with momentum rather than a decision.
Two
Forbid yourself the scroll-up
Drafting and editing use different attention and cannot be interleaved without cost. The moment you go back to improve a line, you have switched modes, and coming back down into the draft takes several minutes you will not notice losing.
When something is wrong behind you, leave a marker and keep going. Write [fix: wrong name] or [check: which hospital] inline and move on. A draft with forty brackets in it is a finished draft; a perfect first chapter is not a book. What to do with them afterwards is the subject of editing a first draft.
Three
Move research out of drafting time entirely
Looking something up is not cheating on your word count, but doing it at 9:40 in the middle of a paragraph is. The browser is a different machine with a different pull, and thirty seconds of checking a date rarely stays thirty seconds.
Bracket it, keep writing, and batch the lookups into a separate block later in the day when writing is over. Most of them turn out not to matter: the scene gets cut, or the detail never needed to be on the page.
Four
Work in short timed blocks, not open afternoons
A four-hour window is not four times a one-hour window. Attention for new prose runs out somewhere between forty and ninety minutes for most people, and the remainder of a long session is usually spent on the rereading column.
Two blocks of fifty minutes with a real break between them outproduce an unbroken afternoon, and they are far easier to schedule, which matters more than it sounds: the writers who finish books are the ones whose sessions fit into the life they actually have.
What professional daily output actually looks like
Comparison is poisonous when the numbers are invented, so here are reported ones. The spread is the point: there is no correct figure, and several of these writers won the same prizes.
2,000
Stephen King's stated daily target, every day including holidays
500
Hemingway's typical morning, counted and logged on a chart
1–2
pages a day was Graham Greene's fixed quota, then he stopped
~250
a page a day finishes a novel in a year with weekends off
Pick the number you can hit on a bad day, not a good one. A quota you miss twice becomes a quota you abandon, and consistency at three hundred words beats heroics at three thousand followed by nine days of nothing.
When the slowness is telling you something
Not all resistance is procrastination. There is a specific kind of stuck that no technique fixes, and it has a recognisable signature: you can write any scene in the book except this one, and you have rewritten its opening four times.
That usually means the scene is wrong rather than unwritten. Often the plot requires a character to do something they would not do, and the part of you that knows them is refusing to type it. Speeding up will not help; changing the scene will.
The distinction is practical. Ordinary slowness is diffuse — every day is a bit sludgy. Structural slowness is local — one chapter, always the same one. If the resistance has moved around the book for three weeks, treat it with sessions and quotas. If it has been parked in the same place, treat it as a block with a cause and go looking for the bad decision two chapters earlier.
Fewer decisions, more sentences
Writer Atlas keeps the book in one place so a session starts with writing rather than with hunting: every chapter in a document tree, search across the whole manuscript when you need to find what you called something, character cards for the details you would otherwise go looking for mid-paragraph, and version history so nothing you cut is ever really gone.
Start writing freeQuestions about writing faster
How many words a day should I write?
Choose the number you can hit on a difficult day. For most people with a job that is somewhere between three hundred and a thousand words. Full-time novelists commonly report between one and three thousand. A page a day, around two hundred and fifty words, finishes a full-length novel inside a year — the figure matters far less than whether you hit it most days.
Does writing fast produce worse prose?
It produces a rougher first draft, which is not the same thing. Final quality is decided in revision, and a fast draft gives revision more to work with while the whole shape of the book is still in your head. The genuine risk is structural: writing quickly without knowing where a chapter is going can bury a wrong turn under twenty thousand words.
Do sprints and timers actually work?
For many writers, yes, because a visible deadline suppresses the urge to edit. Twenty-five to fifty minutes with no scrolling back is the common form. They work badly for material that needs real thought per sentence — a dense argument, a technical explanation — where the timer only adds pressure to a process that is genuinely slow.
Should I use dictation to write faster?
Dictation can triple raw output and it changes the prose: spoken sentences are longer, more repetitive and more conversational, so the cleanup pass is substantial. It suits dialogue, memoir and first-person narration far better than dense exposition. Try it for a week on a scene you already know before deciding.
Is it better to write every day or in long weekend sessions?
Daily wins for most people, because the reload cost is the hidden expense. Returning to a book after six days means rereading to remember where you were, which can eat the first hour of a session. Thirty minutes a day keeps the book warm. If your life only permits weekends, start each session by rereading the last page only and then write forwards.
How do I stop editing while I draft?
Make the fix cheaper to postpone than to perform. A bracketed marker takes two seconds and preserves the thought, where going back to rewrite the sentence costs several minutes of momentum. Some writers also dim the screen or hide the preceding page. The habit takes about two weeks to form and is the single most reliable speed gain available.