How to Write a Novel: Plan, Draft, and Revise Your First Book

1. Test whether your idea can sustain a novel

A strange discovery can produce a strong opening. For a longer book, investigate what acting on that discovery would disturb. You need reasons for the protagonist to keep making consequential decisions after the first problem appears.

Try writing three consequences of Nell's first investigation. An old colleague refuses access to a survey archive. Her sister warns that public accusations could cost them the family house. A former resident supplies a photograph that contradicts the official date. Each development can change whom Nell trusts and what she tries next.

Now imagine the easiest way she could quit. She could file the new measurements and leave. Give her a believable reason to continue, such as finding her own childhood handwriting on a map she does not remember seeing. Avoid solving this problem with a vague statement that she is naturally curious.

If your idea only produces one confrontation, it may suit a short story. That is useful information. You can complete the smaller version, or explore another person's stake in the situation before committing to a book.

2. Decide what kind of reading experience you are making

Choose a provisional genre and audience. Nell's book could be a mystery about the falsified map, a family novel about inherited shame, or a speculative story in which the submerged town has changed. Each choice asks you to develop different scenes.

For this example, we will write a mystery with a family conflict running through it. The reader should be able to follow Nell's investigation, form suspicions, and understand how the evidence supports the eventual answer. Her relationship with her sister will affect the investigation rather than occupy a separate, detachable plot.

Read a few novels close to the kind you want to write. Make notes about where the main problem appears, how information is withheld, and what the ending resolves. Describe the function of a scene in your own words. Do not borrow its distinctive characters or sequence of revelations.

Use a provisional length to estimate the work. We will use 75,000 words for the schedule later in this guide. That is a planning assumption for this imaginary project, not a universal requirement. If you intend to query an agent or enter a competition, check the current requirements for that audience and genre.

The broad question of how to write a book also includes memoir and nonfiction. This guide concerns a novel: a sustained work of fiction. Research, argument, and permissions can create different planning needs for other kinds of books.

3. Build characters who make each other's choices harder

You can learn a great deal about Nell without filling out a long biography. Decide what she wants from the investigation, what she wants from her sister, and what evidence she would be tempted to dismiss. Those answers will influence her behavior in scenes.

These are starting positions. Ruth may eventually decide that keeping the house is less important than knowing what their father did. Vale might believe a false account because someone he trusted gave it to him. Let discoveries change people's positions instead of having them repeat the same argument in every chapter.

In his Reedsy novel guide, Tom Bromley connects a protagonist's external problem with an internal conflict. Here, Nell's effort to establish the facts also tests her belief that her family has always been unfairly blamed. Read the source.

Give each supporting character something to do while Nell is absent. Ruth might meet a buyer. Owen might contact a journalist. Their independent actions let the situation change between Nell's visits and prevent the cast from feeling as though they are waiting for her.

4. Choose a viewpoint and research the scenes that need it

For this version, use close third person through Nell. Readers can inspect the evidence she finds, while Ruth's intentions remain partly uncertain. If you add Ruth's viewpoint, decide what her scenes contribute and which uncertainties they remove.

Try one scene before committing to a complicated viewpoint structure. Nell asking Ruth for the archive key will tell you more about the voice than a page of abstract preferences. If the scene works, write a second one where Nell is alone and must interpret something she has found.

Make a research list from actions you plan to write. How would a surveyor compare maps made in different years? Who controls access to the reservoir? What makes a recently exposed building unsafe? These questions affect the plausibility of Nell's choices. Verify them with suitable primary sources or qualified people before treating the scenes as technically accurate.

Keep a distinction between verified facts and invented arrangements. Our example uses a fictional town and has not established real surveying procedures. Mark uncertain details in the manuscript so an attractive bit of invented procedure does not accidentally become a claim about real practice.

Keep research notes outside the draft. A note can explain a whole system; a scene may only need the reason Nell cannot enter a building today. Include the explanation at the moment the reader needs it to understand her next action.

5. Outline the changes that lead toward an ending

Write a provisional ending before planning every chapter. Nell might prove that the map was altered while learning that her father acted under pressure. She must then decide what to make public, knowing the evidence will implicate someone still alive. You can change this ending as the draft develops.

Work back from that decision. What would she need to learn? Who could prevent her from learning it? Which earlier action would make the final choice more difficult? This gives you a chain of events to test.

These rows are turning points, not six compulsory chapters. A discovery might take three scenes. An aftermath might take one. Expand the outline where you need help and leave room where you want to discover the story through writing.

If outlining makes you stop caring about the book, plan only the next two scenes and keep a sentence about the likely ending nearby. After each writing session, record what actually happened. That running record will help you see the shape you are making.

6. Give the middle new problems to solve

The middle becomes repetitive when the protagonist keeps trying the same approach against the same resistance. Nell asks for a file, receives a refusal, and asks somebody else. If several chapters follow that pattern, change what the previous attempt has made possible or impossible.

You now have two changes: a new piece of information and a damaged relationship. The next chapter can develop either one. A quiet scene in which Nell decides what to tell Ruth can carry weight because the earlier encounter has changed the decision.

For every proposed subplot, ask where it changes the main action. Ruth selling the house could matter because the survey records are stored there, and a sale would force the sisters to sort them. A separate romance that never affects Nell's choices may need a stronger connection or a smaller role.

End chapters where the reader has a reason to continue. Sometimes that is a discovery. Sometimes it is a decision whose consequences we have not seen. You do not need to manufacture a physical threat at every break, especially if the following chapter immediately cancels it.

Keep an unanswered-question list. Next to each question, note when you plan to answer it and what evidence the reader will have seen. This is particularly useful for a mystery, but it also helps track family secrets and promises in other genres.

7. Make a schedule from time you actually have

Choose a few repeatable writing sessions and test them for two weeks. Track the time you spent and what you accomplished. Some sessions will produce new pages; others will solve a problem that lets you continue. Both belong in your estimate.

If you can manage only two sessions a week, use that reality in the calculation. A smaller regular commitment is easier to assess than a daily target you repeatedly miss. Adjust after the trial period rather than treating the first estimate as a test of whether you are a writer.

Before ending a session, leave three lines about what happens next. Note the person in the scene, their immediate aim, and the first action you intend to write. This reduces the number of decisions you must make when you return tired.

Keep dated backups outside the device you write on. Check that you can open an exported copy. A writing routine should protect the work you already have as well as make space for the next chapter.

8. Finish a draft while keeping track of what needs repair

Drafting and revision can overlap. You may need to fix a scene before its consequences make sense. The difficulty is knowing when a repair helps you continue and when polishing the opening has become a way to avoid the unwritten middle.

Keep a repair log alongside the manuscript. If Nell needs to have seen a document in chapter two, note the change and write the later scene as though the document is already there. If the discovery changes her entire motivation, pause long enough to revise the outline. Those problems need different amounts of work.

• Continuity note: Ruth has the archive key in chapter four; explain how Nell obtains it before chapter six.
• Research note: confirm whether the assumed access arrangement is plausible.
• Structure note: the second interview repeats information already established; combine its useful discovery with another scene.
• Sentence note: the description of the exposed street is vague; revisit after the scene order is settled.

Write an ending even if parts of the route remain rough. Completing it reveals which earlier promises you kept, which you forgot, and which no longer belong. A provisional final chapter gives the revision a destination.

9. Revise from the whole book down to the line

After a break long enough to regain some distance, read the draft as a continuous story. Keep a separate note of where your attention drops or the logic breaks. Try to finish that read before spending an afternoon perfecting one paragraph.

Make a reverse outline: one row per existing scene, recording the viewpoint, immediate aim, change, and unresolved question. You are describing the manuscript you wrote, even where it differs from the plan.

Revise the largest dependencies first. Changing who withheld a document could alter six scenes. Correcting those scenes before polishing their sentences saves you from carefully editing material you will later remove.

Then ask readers who enjoy the kind of book you are writing for focused feedback. Ask where they expected a different outcome, whose behavior they did not believe, and which questions remained unanswered. Give them permission to describe their experience without proposing a solution.

Look for repeated observations across readers, but examine the page before accepting a proposed fix. Two readers losing interest at the same interview is worth investigating. Their suggestions to add a chase or a romance may not fit your novel.

10. Decide when this version is ready for its next reader

You can keep discovering possible improvements indefinitely. Set a purpose for the next version. A draft for a critique partner needs a complete arc and clear questions for feedback. A manuscript for submission needs careful editing and the recipient's current formatting requirements.

Read the opening and ending together. Check whether they belong to the same book. Nell's opening promises an investigation shaped by family loyalty. The ending should address both, even if it leaves the sisters' relationship uncomfortable.

Save the manuscript in an editable format and create a separate reading copy. Inspect chapter breaks, missing text, and any notes you meant to remove. Exporting a file is useful, but it does not by itself make the manuscript ready to publish.

Common questions