How to Write a Novel With a Full-Time Job

Most advice on this topic is about discipline, which is the least interesting of the obstacles and the only one you already knew about. Yes: you must sit down. Yes: five hundred words a day finishes a novel in six months.

The reasons people with jobs don't finish novels are more specific than laziness, and each has a mechanical fix.

Problem 1: The re-entry cost

You have forty minutes. You open the file. You spend eleven of them rereading the last two pages to remember where you were, four more deciding that the last two pages are bad, and then the window is half gone and you haven't written anything.

The cost of returning to a novel is far higher than the cost of continuing in one. A writer with unbroken afternoons pays it once a day. You pay it every session, and your sessions are short, so it eats a proportionally enormous share.

The fix is to end deliberately. Hemingway's version: stop when you know what happens next. Stop mid-scene, mid-argument, even mid-sentence — somewhere the next move is obvious. Then leave yourself one line of instruction at the bottom of the file: "Next: she lies about the car, he lets her."

Tomorrow you open the file and you are already writing. You have converted eleven minutes of orientation into zero.

Problem 2: The forgetting

A full-time writer holds the novel in working memory. You cannot. Between Tuesday and Saturday, forty hours of other people's problems will pass through your head, and what leaves is exactly the thing you needed: the innkeeper's name, the fact that the sister already knows, what you had decided the ending meant.

Then you write around the gap. You avoid the scene with the minor character because you can't remember what you established about him. You skip the confrontation because you're no longer sure who knows what. The novel silently reshapes itself to fit what you can still remember, which is a terrible principle of composition.

The fix is to externalise the state. Not diligent note-taking in a separate document — you will not do that at 6 a.m. — but a reference that lives beside the draft and that you can consult in seconds. Who these people are, what they know, what you promised. The test is whether looking something up costs less than avoiding the scene. If it doesn't, you'll avoid the scene.

Problem 3: The wrong metric

Word count is a bad target for people with fragmented time, because word count rewards the sessions where you barrel forward and punishes the ones where the right move was to think. Some days the correct output is four hundred words. Some days it's deciding that the second act is wrong, and writing nothing.

But don't replace it with nothing. Untracked writing decays, because you cannot tell the difference between a slow month and a stalled one until it has become a stalled year.

Track two things: did you open the file, and did the book move. A streak of days-touched is a better predictor of finishing than any word count, because the enemy is not slowness. It's the gap that becomes a month, and then a drawer.

Problem 4: Revision needs a different kind of time

Drafting fits in forty-minute pieces. Structural revision does not: it requires holding the whole shape at once, and forty minutes is barely enough to load it.

So don't attempt them in the same units. Draft on weekday mornings, in small pieces, forward only. Save the structural work — the moving of chapters, the diagnosis of the sagging middle — for the one long stretch you get every week or two, and prepare for it by making the shape visible rather than something you must reconstruct from memory each time.

An outline you can see is an outline you can think about in the fragments of attention you have. A ninety-thousand-word document is not.

What actually finishes books

Not intensity. Almost every finished novel written around a job was produced by someone boring: a fixed slot, a low target, an obvious next move waiting in the file, and a reference document that carried the memory the writer couldn't.

The romantic version — waiting for the long free summer, the retreat, the moment you feel ready — has produced fewer novels than the person who writes three hundred mediocre words before work, five days a week, for two years.

Six hundred days. Three hundred words. That's a novel, and it's longer than most.

Narrale is built for the fragmented kind of writing: a distraction-free page that opens where you left it, a Story Bible that remembers what you can't, daily goals and streaks that measure momentum rather than volume, and a planning canvas for the weekends when the whole shape has to be visible at once.

Related: Writer's block is a symptom · Writing faster with AI · How to outline a novel