Sagoræ · Blog

What Makes a Good First Draft of a Novel? What It Needs and What It Can Get Wrong

Published 9 November 2026 · 9 min read · Novel Craft

What Makes a Good First Draft of a Novel? What It Needs and What It Can Get Wrong

A good first draft of a novel is not a well-written text. It is a finished one: every scene from the opening to the last page, something you can read in one sitting and follow as a story. It needs an ending, a main character who wants something, and scenes that lead into each other, however clumsily. Almost everything else is allowed to fail: the prose, the repetitions, the descriptions, chapter lengths, character names, and a fair chunk of the middle.

Below is where that line actually falls, with a table to keep at hand, a before and after excerpt, an eight-question test and an end-of-draft note you can copy.

Why writers misjudge their first draft

Most new novelists grade their first draft against a published book. They open chapter 3, hold it up next to the novel on their nightstand, and decide they can't do this. The comparison is rigged. The nightstand novel has been rewritten, cut, edited by a professional and proofread twice. You are comparing your scaffolding to a house people live in.

The opposite mistake is quieter and does more damage: believing a first draft needs nothing at all, that you just pile up words. That is how you reach 60,000 words with no ending, three open plotlines and a heroine who changed jobs in chapter 11. That isn't a first draft. It's a heap of material, and revision has nothing to grab.

Between the two sits a simple definition. A first draft exists to be revised. It is good if it gives the next version something to fix. It is bad if it gives revision nothing to hold on to.

What a first draft has to contain

Five things. If they are there, your draft is good, whatever the sentences look like.

A written ending. Not a perfect ending, not even the right one: an ending. Until the last scene exists, you can't know what the opening has to set up, and revision goes round in circles. A lot of writers stall just before the final scene, often because they're afraid of fumbling it. Fumble it. It will be rewritten anyway.

A character who wants something. At least your protagonist needs a readable desire, even if it's badly phrased. "Molly wants to know who keeps writing to a dead man" is enough. A draft where nobody wants anything has no spine.

Scenes, not summaries. First drafts drift into summary ("over the following weeks they got to know each other"). A few summarized stretches are fine. Most of the book should be played out: people in a place, talking or doing something. A wobbly scene can be rewritten. A summary has to be turned into a scene first.

A record of your decisions. When you change something midway (a character's age, the order of two reveals), you don't go back and fix earlier chapters. You write it down. That's what lets you keep moving without losing the thread.

Enough material. A draft that comes in short is harder to revise than one that runs long. Cutting is precise work; adding fifty pages that don't exist means writing a second first draft. Better to finish at 95,000 words aiming for 80,000 than the other way round.

What it is allowed to get wrong

Anything that revision deals with, which is a lot.

ElementIn the first draftWhen to deal with it
The endingMust exist, even if it's wrongRewritten in the structure pass
The protagonist's desireMust be readableSharpened in the character pass
Chapter orderCan be wrongStructure pass
The middleCan sagStructure pass
Secondary charactersCan be vague or surplusCharacter pass
NamesCan be placeholdersAny time, with one find-and-replace
DescriptionsCan be missing or bloatedScene pass
Style and sentence rhythmCan be flatSentence pass
Repetitions and typosAs many as it takesLast pass, or a proofreader
Facts to checkLeft in square bracketsAfter the structure is settled

The logic is the one behind editing a first draft in four passes: heavy problems first, fine ones last. Polishing a sentence in a chapter you will cut next month is wasted work, so the first draft has no reason to polish what revision is going to take apart.

Square brackets: the trick that saves a first draft

The most useful first-draft tool is a key on your keyboard. Whenever you get stuck on a detail that doesn't stop the scene from moving, open a bracket, write what's missing, and keep going.

Here is an invented passage written by someone who stops at every doubt. It took an evening:

Molly parked the red van outside the village hall. Thornside sat at 900 feet (or was it 1,200? need to look up the height of villages in that valley) and smelled of woodsmoke. She took out Mr Fairfax's post. Fairfax, no, too posh. She took out the post.

The same passage written with brackets, in ten minutes:

Molly parked the red van outside the village hall. Thornside [HEIGHT?] smelled of woodsmoke. She took out the post for [DEAD MAN'S NAME, not Fairfax]. Three letters, the same slanted handwriting as last week. She turned them over. No return address. [Is Molly allowed to open them? Check Royal Mail rules.]

The second version is messier and far better, because it moved: the slanted handwriting, the missing return address, the question of whether she can open them. Those are plot. The first version circled a height and a surname.

Always use the same bracket format (capitals, say) so you can find every one with a single search at the end. And set yourself one rule: nothing gets checked during a writing session. Checking gets its own session.

The change log: change your mind without going back

In chapter 14 you realize that Dan, the local police officer, shouldn't be Molly's childhood friend but a newcomer. The natural reflex is to go back to chapters 2, 5 and 9 and fix them. That's how first drafts die. You reread, tweak a sentence, then another, and three weeks later you're still in chapter 6.

Do the opposite. Write on as if the change had always been true, and log it in a separate file:

ChapterWhat changesFix in
14Dan moved to the valley this year and doesn't know MollyCh. 2, 5, 9
17The March letter is typed, not handwrittenCh. 3, 8
21The postmaster knew about the dead man all alongAll his scenes

That table becomes page one of your revision. It beats a reread, because it tells you exactly where to look. If you keep character sheets, the change goes there too: the sheet holds the current truth, and the manuscript catches up later. The routine is laid out in tracking characters, places and timeline.

When is a first draft finished?

You'll never be satisfied with it, so that's not the test. It's finished when it meets the five conditions above. Answer these honestly, yes or no.

  1. Is the last scene written?
  2. Can you say in one sentence what your protagonist wants?
  3. Does someone or something get in the way of that before the midpoint?
  4. Can you sum up each chapter in one line without writing "not much happens"?
  5. Are the changes you made along the way written down somewhere?
  6. Are the gaps (facts, names, missing scenes) flagged with a marker you can search for?
  7. Do you have roughly as many words as your target length, or more?
  8. Could you hand it to someone you trust and say "it's rough, but it's a whole story"?

Reading your answers. Eight yeses: the draft is done, put it away. A no on question 1, 2 or 4: it isn't done, and that's where to go back to, not the prose. A no on 5 or 6: it's done, but spend one evening listing the changes and gaps before you shelve it, or you'll have forgotten them in a month. A no on 7: look for summarized passages that could become scenes. A no on 8 alone: that's probably nerves, not the text. The finer signals come later, on the read-through, and they're covered in is my novel any good?

If the outline itself feels shaky, now is the time to stress-test it, before you touch a single sentence. Here is the kind of question a structural read asks of a finished draft:

D Daedalus · Story architect Pre-recorded demo

Daedalus stress-tests the outline of "What the Sea Gives Back"

You

Here is my three-act outline. I am stuck in the middle. It feels like nothing happens between chapters 9 and 16.

Act I (ch. 1-8): Maud takes over her father's cannery. Act II (ch. 9-22): she discovers the debts, looks for a buyer, meets Jonas. Act III (ch. 23-28): the cannery burns down, Maud decides to stay.

Daedalus

Your feeling is right, and the outline shows why. From chapter 9 to chapter 22, Maud looks for a buyer. Looking is not an event: it is a state. Nothing stops her from looking for thirteen chapters.

Three questions, and no answers on your behalf:

  • Where is the midpoint? Something around chapter 14 has to make going back impossible. A buyer who signs, then pulls out? A discovery about the father?
  • What does Jonas want? He appears in Act II with no goal of his own. A character who wants nothing does not create conflict, he creates a pause.
  • Why does the cannery burn? If it is an accident, the climax comes from outside. If it is the consequence of a choice Maud makes in Act II, the whole middle tightens at once.

You

What if Jonas were the buyer, and he wanted to tear the cannery down to build flats?

Daedalus

Then your Act II has just found its spine: Maud needs the money of the man who wants to destroy what she is trying to save. Your midpoint almost writes itself (she signs, or refuses to sign), and the fire in chapter 23 changes meaning depending on who holds the match.

I am not updating anything: it is your outline. Add the line "Jonas = buyer" to chapters 9 to 16 and look at what each chapter now has to contain.

Pre-recorded simulation. Daedalus never writes: it questions the structure of the novel and leaves the decisions to the author.

The end-of-draft note, ready to copy

On the day you type the last word, before you close the file, fill this in. It takes twenty minutes and saves dozens of hours of revision, because four weeks from now you won't remember what you know today.

End-of-first-draft note

  • Date:
  • Word count:
  • The novel in one sentence (what the character wants, what stops them):
  • The scene I'm most sure of:
  • The scene I'm least sure of:
  • The chapter where I nearly quit, and why:
  • What the ending changed about how I see the opening:
  • The three heaviest changes logged along the way:
  • The character who surprised me:
  • The character who may not earn their place:
  • Brackets still open:
  • Planned restart date (at least four weeks away):

The line about the ending is the one that matters most. Writing the ending is very often when you find out what the book was really about. It tends to be a little different from what you thought in chapter 1, and that discovery steers the whole rewrite.

What to do next

Nothing, for a month. Then read it in one sitting without fixing anything, and start on structure. The full schedule, with timings for 80,000 words, is in the four-pass method. If you're still stuck in the middle and drowning, start with the triage guide for 30,000 words and where writers actually quit: it's almost always the same places.

In Sagoræ, the "Lire comme un livre" view (read as a book; the interface is in French for now) opens your manuscript laid out full screen, with none of the writing tools around it. It's a good way to do that first straight-through read with a bit of distance, as if someone else had written it.

Frequently asked questions

What is a first draft in writing? It's the first complete version of a text, written start to finish without trying to perfect it. For a novel, it's complete when the last scene exists and the story can be followed from beginning to end, even clumsily.

Should I edit my first draft as I write? No, with one exception: rereading the last few lines from the day before to pick up the thread. Everything else waits for the end. Changes go in a change log, and doubts go in square brackets.

How long should a first draft rest? Four weeks is a sensible minimum. You need to read what you wrote, not what you meant to write, and a few days isn't enough distance for that.

My first draft is terrible. Is that normal? It's normal for it to be badly written. It's less normal for it to have no ending or for nobody in it to want anything. Judge it on the five conditions in this article, not on how the sentences sound.

How many drafts does a novel need? There's no fixed number. A well-ordered revision, from structure down to sentences, matters more than the count. What counts is that each pass fixes one level of problem without undoing the one before.

If you want to write your first draft with your character sheets, outline and notes in one place, then read it back as a book when it's done, try Sagoræ for free.

✒️ Read next: How to Edit Your First Draft: A 4-Pass Method From Structure to Commas · How to Finish a Novel: Where Writers Actually Quit, and How to Get Past It · I've Written 30,000 Words and I'm Lost: A Triage Guide · Is My Novel Any Good? 5 Measurable Signals Before You Show It to Anyone · Story Bible for a Novel: Track Characters, Places and a Dual Timeline Without a Spreadsheet

Write your novel in a real workshop: a focused editor, characters, story structure and AI coaches that guide you without writing for you. Start for free, no card required.