How to write a first draft

How do you write a first draft? The question is not really about writing; it is about finishing. A first draft is the version of the story whose only job is to exist, complete, from first scene to last, and every practice that matters during drafting exists to protect that completion. This guide covers the draft's job, the momentum practices craft sources name, the famous permission framing, how to tell when the draft is done, and the handoff into revision. How to write a novel covers the whole climb; this page is the middle of it.

The draft's job is existence, not quality

The argument is old, and the sources state it plainly. Writer's Digest, in Martha Alderson's seven reasons for drafting straight through, puts the core insight near the top: "Until you write the end, you do not have a clear grasp of what comes earlier." The draft is where you find out what the story is. You cannot know what belongs in the opening until the ending exists, so polishing chapter one before chapter forty is written is polishing a page whose job is not yet knowable.

That is also why the draft is allowed to be bad, a point the same article makes without blinking: one of the benefits of a "truly awful, lousy, no good first draft" is "that it can only get better from there." Craft's most famous version of this permission is the "shitty first draft" framing usually credited to Anne Lamott's Bird by Bird (1994). A sourcing caveat worth keeping: the book itself was not one of this guide's fetched sources, so the attribution stands as widespread craft consensus rather than a verified quotation. The permission is the point either way: the draft's prose does not have to survive revision, and most of it will not.

The cost of ignoring this is concrete. Alderson notes that most manuscripts lose "thirty-five to one hundred pages" off the front before they are finished, and her warning is the one worth taping to the monitor: "The less time you devote to making every word perfect in the first couple of drafts, the less painful future cuts and revisions will be." Perfecting words you will later delete is the most expensive writing there is. The drafting posture she recommends is tolerating "the chaos of ugly prose, gaps, and missteps in the early drafts."

Momentum practices

• Draft forward only. The organizing rule of the Writer's Digest piece: write the entire draft before going back to the beginning. Restarting chapter one repeatedly feels productive and quietly prevents the one thing that teaches you the book, which is reaching its end.
• Keep moving, in any direction. Reedsy's writing coach Tom Witcomb gives the momentum rule a practical shape: "The key to finishing a book is to keep doing something (writing words, not perfecting your file-naming structure), even when it gets tough." His personal version allows lateral movement: "when my interest starts to wander, I can break away from it and move on to another scene." Forward only does not have to mean linear only; it means never backward into polishing.
• Defer the prose. Witcomb drafts his scenes while "not worried about the writing yet": story first, sentences later. If a sentence fights you, write the ugly version and flag it.
• Placeholder everything you would stop to research. Craft consensus, and standard practice: when the draft needs a fact, a name or a description you do not have, drop a bracketed placeholder ([CHECK: harbor depth], [name], [describe her cloak later]) and keep writing. The convention exists because research is the most convincing form of procrastination available while drafting. The placeholders also become the revision to-do list; the ones about people and places are exactly what how to describe a character picks up later.
• Shrink the daily goal until it is unmissable. Witcomb's advice for consistency: "Aim for 500 instead and you'll be much more likely to consistently hit that goal." A target you hit every day builds the chain; a heroic target you hit every third day breaks it.

Plotter, pantser, and the draft

Whether you outline first is a separate decision, covered in plotter vs pantser, and both kinds of writer end up in the same place. Pantsers discover the story in the draft; plotters discover how the story diverged from the plan, also in the draft. Either way the draft's job does not change. If you did outline, how to outline a novel makes the pairing rule explicit: when the draft deviates, update the outline the same day, and the plan stays a working map. Drafting time is also the cheapest time to keep a story bible: a few lines logging what the day's pages made true costs nothing now and saves a continuity hunt at revision.

The serial caveat

A web serial narrows the luxury this guide describes, because a serial publishes its first draft live: the polish pass readers get is per chapter, not per book. The practical adjustment is to raise the quality floor slightly at the chapter level, which is its own subject in how to write a chapter, and to lean harder on the outline and story bible, since a serial cannot revise what has already published. Momentum matters more, not less: a release schedule does not pause for writer's block.

When is a draft done

Done is a structural fact, not a quality judgment. The draft is done when the story exists end to end: the final confrontation is written as a scene rather than promised in a note, the last line exists, and the document says what happens. Not "the prose is good." Witcomb's line about the moment of finishing is bracing: "Only the truly insane would get to the end of their first draft and think they were done." Done means the material for revision finally exists, all of it.

Two tests help. Can you state the ending without hedging, including what it costs the protagonist? If any stretch of the middle is still a summary bracket, the draft is not done. And can you write the one-sentence version of the whole story now, when you could not in chapter one? That sentence is the draft's real product: it is what the next version of the book gets built around.

The handoff into revision

The transition is a gap, not a slide. Witcomb tells clients to "put the manuscript in the bottom drawer" and take a break from it, the same cooling-off practice the catalog's editing guide builds on: how to edit your novel opens with that gap and walks the passes that follow, story, then scene, then line. Do not reread the draft on the way to the drawer; the reread belongs on the editing side of the gap, with fresh eyes.

The whole arc, then: permission to write badly, momentum practices that keep you moving forward, a placeholder for everything that would stop you, an ending written rather than intended, and a deliberate pause before the first pass. The draft is not the book. It is the first version of the story that can actually become one.