How to Write a Novel

Short answer

• To write a novel, start with a clear premise: a character, a goal or problem, and stakes that matter to that character. Test the idea by listing scenes it could produce. If the idea can generate many scenes and a change in the character, it may be strong enough for a novel. Next, decide how much to plan. Some writers outline in detail, and others discover the story as they draft. Either approach works if you keep track of the main story question, the turning points, and the character change. Then set a regular writing routine, draft the whole story without trying to make it perfect, rest it, and revise in passes: first for structure, then for scenes and character consistency, and finally for language and errors. A first draft is a working version, not a finished book. Revision is where the novel takes its final shape.

By Emma Hudson

Writing a novel can feel very large at the start. The finished book may be tens of thousands of words long, and the path from a first idea to a revised manuscript has many stages. It helps to break that path into smaller decisions: find an idea that can sustain a long story, shape it into a premise, build characters who want something and face a real obstacle, plan enough structure to keep moving, and write a first draft that you can revise later. This guide walks through each stage in order, with routines you can adapt to your schedule, genre, and working style.

1. Find and test an idea that can carry a novel

A novel idea does not need to be grand. It needs to be durable, so it keeps producing questions, scenes, and complications after the first excitement fades. A striking image, a question you cannot stop turning over, a character voice, a place, or a conflict can all be enough to start, provided the idea can grow. Try writing the idea in one or two sentences, adding a character, a problem, and a stake. For example: a retired ferry captain returns to an island town to sell her family home, and discovers a letter that changes what she believed about her brother’s disappearance. A theme alone, such as grief or ambition, is not yet a story. A story needs someone who acts, wants, or resists something.

Test the idea with a scene list. Set a timer for ten minutes and list every scene you can imagine, using only a few words for each scene. If you can list many scenes easily, and those scenes suggest rising complications rather than repetition, the idea has room to grow. If you struggle to reach even a handful, add a stronger problem, a second character with a conflicting goal, a higher stake, or a larger character change. Also ask what would happen if the main character did nothing. If nothing important would change, the stakes may be too low. The Purdue Online Writing Lab suggests invention questions at the start, including what story you think others should hear and what makes your idea different. Use those prompts to clarify what draws you to the idea, then keep an idea notebook and return to it over days or weeks.

2. Turn the idea into a premise and a story question

A premise is a compact statement of the story. It usually names the main character, the situation, the goal or problem, and the obstacle or stake. A premise is not a summary of the whole plot. It is a tool for making decisions while you draft. When you are unsure whether a scene belongs, ask whether it moves the character closer to or farther from the goal, deepens the problem, or reveals character in a way that matters later. Write your premise on a single page or card and keep it visible. You may change it as the story develops, and a premise that changes because you discovered a stronger conflict is progress, not failure.

A story question is closely related. It is the question the reader carries through the book: Will the captain find out what happened to her brother, and what will she do with that knowledge? The question should be specific enough to create curiosity, but open enough to sustain many scenes. If it can be answered in one scene, the novel needs a larger question or a series of linked questions. Finally, imagine the ending in broad terms. You do not need every detail, but you should sense what change or resolution the story is moving toward. An ending does not need to be happy, but it should feel connected to the problem set up at the start.

3. Build characters readers can follow

Readers follow characters, not just plots. A useful character has a want, a need, a flaw or fear, and a capacity to change. The want is the visible goal: to solve a case, to win a competition, to return home, or to protect a family member. The need is often less visible: to accept help, to tell the truth, to forgive, or to take responsibility. The tension between want and need can drive many scenes. A character who wants safety but needs to take a risk, for example, will face choices that reveal both sides. Give the main character strengths as well as limits, and choose background details that affect decisions, relationships, or the plot, rather than filling the draft with every detail you invent.

Secondary characters should have their own goals, even if those goals are small. A friend who wants the main character to stay safe may oppose a risky plan for a good reason. An opponent who wants the same prize for a different reason is more interesting than an opponent who is simply cruel. The Purdue Online Writing Lab’s invention questions for fictional characters are a practical starting point: the character’s age, greatest wish, greatest fear, hobbies, home, speech, appearance, self view, relationships, job, and situation just before the story begins. Answer the questions that affect the story, and return to the rest when a scene needs more depth.

4. Understand structure, scenes, and chapters

Structure is the shape of the story over time. Many novels use a broad three part shape: a setup, a middle in which complications grow, and an ending in which the main problem reaches a resolution or a decisive change. In the setup, the reader meets the main character in a normal situation, learns what matters to that character, and encounters the event that disrupts that normal life. In the middle, the character tries to solve the problem, meets obstacles, and faces rising stakes, often with a turning point near the middle that changes direction, such as a revelation, a loss, or a decision that cannot be undone. In the final part, the character faces the main problem directly, makes a decisive choice, and the story shows the result.

Think in terms of turning points and causal connection. A turning point is a moment after which the story cannot go back to how it was. Does each major event lead to the next event, or does the story simply jump from one incident to another? When events cause later events, the story feels purposeful. A novel is built from scenes, each usually in a specific place and time, following a particular point of view, with a goal, an obstacle, and an outcome that changes the situation. Show important turning points, choices, and revelations as scenes, and summarize travel, routine days, and background that only bridge scenes. Chapters are a pacing tool: short chapters can speed up a tense section, longer chapters can slow the pace for reflection, and a chapter can end at a decision, a revelation, or a question that makes the reader want to continue.

5. Choose point of view, tense, setting, and research

Point of view decides whose perspective the reader experiences. In first person, the narrator is a character who tells the story as “I,” which creates intimacy but limits the reader to what that character knows, sees, or guesses. In third person limited, the narrator follows one character closely in each scene, using “he,” “she,” or “they,” and can report that character’s thoughts without revealing other characters’ inner thoughts directly. In third person omniscient, the narrator can move between characters, but frequent shifts can distance the reader if they are unclear. Tense is a separate choice: past tense is common, while present tense can feel immediate but requires care with reflection and backstory. What matters most is consistency within a scene and a clear signal for any shift, such as a chapter break.

Setting is more than a backdrop. It can create pressure, limit choices, and reveal character. A small town with one ferry a day creates different problems from a large city with many exits. Research helps fiction feel believable, even in invented worlds. If your novel is set in a real place or period, check layouts, jobs, technology, laws, customs, and language for that time and place. Keep a separate research file, and use placeholders for uncertain details while drafting, so research does not stall writing. For invented worlds, decide on rules for any invented system, such as magic or future technology, and apply them consistently, with a short world guide for place names, timelines, and key rules to prevent contradictions.

6. Decide how much to outline

Some writers outline in detail before drafting. Others write by discovery, starting with a character and a situation and learning the story as they go. Most writers fall somewhere between these approaches. A detailed outline can help you see the whole shape, avoid a stalled middle, and write faster. Discovery writing can produce surprises, strong voice, and unexpected connections. Both approaches require revision. An outline does not prevent revision, and discovery writing does not mean writing without any plan.

A simple outline can be a list of scenes in order, a chapter by chapter summary, or a set of turning points with notes on character change. Start with the main story question, the setup, the major turning points, and the broad ending, then fill in the middle with scenes that raise stakes, test the character, and reveal information. Leave room for change: if a character makes an unexpected choice during drafting, update the outline rather than forcing the draft back to an old plan that no longer fits. If you feel lost, pause and outline only the next few chapters. A short term plan can restore momentum without locking you into a structure you have not yet tested.

7. Set a routine and write a realistic first draft

A novel is finished through regular work, not through occasional bursts alone. Choose a routine that fits your life and protect it as much as you can. The best time is the time you can repeat, whether early morning, late night, or short sessions during lunch breaks. Set a target that is challenging but realistic: a word count, a time block, or a scene per session, or a combination, such as writing for 45 minutes or finishing one scene, whichever comes first. Track your progress in a simple log, with date, words written, and a short note about what happens next, so restarting the next day is easier. Silence notifications during writing time, and decide in advance how you will handle missed days: return the next day, rather than trying to make up a large backlog in one exhausting sitting.

A first draft is a complete version of the story from beginning to end, written to discover what the story is and how it works. It is not a polished manuscript. It will contain weak scenes, repeated ideas, inconsistent details, and dialogue that needs tightening. That is expected, because the purpose of a first draft is to have material to revise. Many guides describe a typical novel length as about 80,000 to 90,000 words for adult fiction, with variation by genre: fantasy often runs longer, while some romance and young adult novels run shorter. These are general patterns, not rules. Judge a first draft by whether the main story is complete, not by word count alone. When you finish, rest the manuscript for days or weeks, then read it as a reader first, noting where you were confused, bored, surprised, or moved. Those notes become your revision plan.

8. Write dialogue, description, and work through blocks

Dialogue should reveal character, advance the story, or create tension. It should not merely repeat information the reader already knows, or deliver long explanations that no real person would say in that situation. Listen for voice: word choice, sentence length, habits, and what a character avoids saying. Two characters should not sound identical unless that similarity is intentional. Use dialogue tags sparingly, since “said” is usually clear and unobtrusive, and use action beats, such as a character pausing or looking away, to show emotion. Description should be selective: choose details that establish place, mood, or character and that will matter later. A single telling detail, such as a chipped mug a character refuses to replace, can reveal history more effectively than a full inventory. Show key moments as scenes, and summarize background and transitions that the reader needs but does not need to experience in full.

Most novelists face periods when writing feels slow or unappealing. If you do not know what happens next, outline the next three scenes, write a plain summary of the scene before drafting it, or switch to a scene you do understand. If you are bored with a scene, check whether it has a clear goal and obstacle, and add a complication, a conflicting goal, or a choice with a cost. If perfectionism is the problem, set a small target, such as ten minutes of writing, and stop when the timer ends. Keep a “next steps” note at the end of each session, and archive discarded scenes in a separate file rather than deleting them, since a line or idea from a cut scene may fit later.

9. Revise in passes, from large to small

Revision works best in passes, each with a different focus. Start with structural revision. Read for the whole story: Does the premise hold? Does each major event cause the next? Does the character change in a way that feels earned? Are there scenes that repeat the same point, or missing scenes that would explain a jump? At this stage, you may move, cut, add, or rewrite whole scenes. This can feel disruptive, but it is more efficient than polishing language in a scene that will be removed. Next, revise at the scene level. Check each scene for a goal, an obstacle, and a change. Cut filler dialogue, tighten description, and make sure the point of view stays clear. Check timelines, place names, and invented rules, and keep a continuity file for names, ages, dates, locations, and key facts.

Then revise for language. Read sentences aloud when possible, since listening can reveal awkward rhythm, repeated words, unclear pronouns, and overly long sentences. Vary sentence length for pacing, and cut unnecessary words while keeping voice and rhythm. Finally, proofread for spelling, grammar, punctuation, and formatting. Proofreading is the last pass, not the first. It is better to proofread a structurally sound manuscript than to proofread early and then rewrite large sections. Most novels need at least these three kinds of passes, and some need more rounds, especially if feedback reveals a structural problem that affects many scenes.

10. Get feedback and protect original work

Beta readers are early readers who read a draft and tell you how it works as a story. Choose readers who read your genre, can be honest, and can describe their experience rather than only praising or rewriting in their own style. Give them specific questions: Where did you feel confused? Where did you lose interest? Which character did you care about most, and why? Did the ending feel earned? Early feedback should focus on story, character, and structure, while later feedback can focus on scenes and language. When you receive conflicting advice, return to your premise and story question, and remember that you decide which suggestions fit the novel you intend to write. If several readers stumble at the same place, that place likely needs revision, even if they suggest different fixes. Share with people you trust, keep dated copies, and avoid posting a full unpublished manuscript publicly if you later plan to control first publication rights, since different publishers and platforms have different rules.

Your novel should be your own creative expression. You can be inspired by other books and learn from their techniques, but you should not copy another author’s sentences, scenes, characters, or distinctive expression and present them as your own. When you research, take notes in your own words, and mark direct quotations clearly with their source. If you use a song lyric, a long quotation, or another copyrighted work, check whether you need permission, since rules vary by type of work, amount used, and jurisdiction. The U.S. Copyright Office explains that copyright protects original works of authorship, including literary works, and that protection exists from the moment an original work is fixed in a tangible medium, such as when ideas are written down or typed. Copyright does not protect names, titles, short phrases, ideas, methods, facts, or systems. It protects specific creative expression. Registration creates a public record of ownership and can provide additional legal benefits, and it requires an application, a filing fee, and a copy of the work, called a deposit. This article gives general information only, not legal advice. Back up your manuscript in more than one place, with version names showing date and revision stage.

11. Decide what comes after revision

After a full revision, you have several options, and none of them needs to be rushed. You can rest the manuscript again, seek another round of feedback, revise once more, or start a new project while the novel sits. Some writers revise several times before they consider a novel ready to share more widely. Others decide that a first novel was a learning project, and they apply what they learned to a second novel. Both outcomes are legitimate. A first novel that teaches you how to finish a long story has value, even if it is not the book you eventually publish.

If you later explore publication, learn about the options in general terms before committing to one path. Traditional publication, self-publishing, and sharing work in other formats each have different processes, timelines, costs, and requirements. Research those options from reputable publishing and author organizations, read current guidance, and avoid any service or offer that guarantees sales, rankings, or publication success. No honest guide can guarantee an outcome for a specific manuscript, because readers, markets, and editorial decisions vary. Focus on what you can control: a complete, revised manuscript, a clear understanding of your genre and audience, and a careful, informed approach to next steps. Above all, keep the long view. Each completed draft teaches you more about ideas, characters, structure, and revision than an unfinished perfect opening. Finish the draft, revise with patience, and let the novel become what it can be through steady work.

FAQ

How long does it take to write a novel?

It varies widely. Some writers finish a first draft in a few months of regular work, while others take a year or more around jobs, study, caregiving, or other commitments. A steady routine matters more than speed.

How long should a novel be?

It depends on genre and audience. A common general pattern for adult fiction is about 80,000 to 90,000 words, with fantasy often longer and some romance and young adult novels shorter. Tell the complete story, then check during revision whether the length serves the story or needs tightening.

Do I need an outline to write a novel?

No, but you need some way to track the story. Some writers use a detailed outline, others use a list of turning points, and others discover the story as they draft and outline later. If you feel lost, outline the next few chapters. The method should serve the draft, not control it.

How do I start writing a novel if I only have a vague idea?

Write the idea in one or two sentences, adding a character, a problem, and a stake. Then list scenes the idea could produce, and ask what the character wants, what stands in the way, and what would change if the character did nothing. If those answers suggest many scenes and a character change, you have a workable starting point.

What makes a good novel character?

A good character usually has a clear want, a deeper need, a flaw or fear, and room to change. The character should make choices that have consequences, rather than only reacting to events. Secondary characters should also have their own reasons to act.

How do I write a first draft without getting stuck on perfection?

Treat the first draft as material for revision. Use placeholders for uncertain details, write scene summaries before difficult scenes, and keep moving through the story from beginning to end. Set a small daily target and a next steps note for tomorrow.

How many times should I revise a novel?

There is no fixed number. Most novels need at least a structural revision, a scene level revision, and a language and proofreading pass. Revise until the story, character change, and pacing feel coherent, and further changes are making the manuscript different rather than better.

Should I show my first draft to other people?

It can help, but choose the timing and readers carefully. Many writers prefer to complete a draft and do a first structural revision before sharing, so feedback focuses on the story rather than on known rough patches. Ask specific questions about confusion, pacing, character, and ending.

Sources & references

• Purdue University, Purdue Online Writing Lab. Invention for Creative Writing. https://OWL.Purdue.edu/owl/subject_specific_writing/creative_writing/writers/invention/invention_for_creative_writing.html
• Creative Writing Now. How to Write a Novel. http://www.creative-writing-now.com/how-to-write-a-novel.html
• U.S. Copyright Office. What Writers Should Know about Copyright. https://www.copyright.gov/engage/writers/?loclr=blogcop