How to Outline a Novel

Before choosing a method, know what you are trying to buy with an outline. The point is not to plan every sentence. It is to reduce the number of expensive wrong turns, the kind that cost you forty pages and a month before you notice the road went nowhere. A novel is too long to hold in your head at once, and an outline is where you keep the parts you are not currently writing.

Writers fall somewhere on a spectrum from heavy planning to almost none, and the famous labels (plotter, pantser, and the middle ground sometimes called plantser) matter less than a plain question. How much do you need to see before the next scene feels safe to write? Some people need the whole shape. Others need only the next two scenes and a vague sense of the ending. Neither is more serious. The methods below run from light to heavy, and you can mix them, using loose cards for the body of the book and a tight beat sheet only for the climax you are nervous about.

One rule holds across all of them. The outline serves the draft, not the reverse. The moment it starts forbidding a better idea, it has stopped doing its job and you should change it.

The smallest useful outline is a single clear sentence of premise. Name the character, the want, and the obstacle. A clerk in an invented port town wants to clear himself of a theft he did not commit before the magistrate arrives. That sentence is not a plot, but it tells you what every scene has to push toward, and it catches the most common structural failure early, a story with no engine.

One step up is a set of signposts: three to six events you are fairly sure the book passes through, written as a bare list. The first betrayal. The flight from the town. The moment he decides to go back. The confrontation at the docks. The cost of winning. You are not claiming to know how he gets from one to the next, only that these are the load-bearing moments. Signposts give direction without dictating the route, and they reorder easily when the draft argues with you.

A one-page sketch fills the gaps loosely, a paragraph per act or per quarter of the book, in plain summary. That is enough for many writers to start drafting with confidence and still leave most of the discovery for the page. If you freeze when over-planned, stop here and begin.

Index cards are the most flexible mid-weight method, which is why they have outlasted every tool meant to replace them. One card per scene. On each, write the few things that make a scene a scene: whose point of view it is, what that character wants, what goes wrong, and how the situation has changed by the end. Keep it to a few lines. A card you can read at a glance is a card you can rearrange.

The power of cards is physical reordering. Spread them on a table or a wall, and the shape of the book becomes visible in a way a linear document hides. You can see three quiet scenes in a row with no pressure between them, two cards that are secretly the same scene and should merge, or a midpoint that arrives too late. You move the cards, not the prose, which is cheap. A digital binder of short scene documents does the same work, and reordering a list of scenes is far less painful than reordering thousands of finished words.

A scene list is the same idea written down the page instead of across a table: a numbered list where each line is one scene, often with a goal-and-disaster shorthand. Number nine, he picks the lock, takes the wrong letter, and is now named as the thief. A good scene list is the densest map most novels need. You can hold the whole book on two pages and still see every joint.

Beat sheets are the heaviest common method, and they work by giving you named slots to fill. The three-act structure is the oldest version: a setup that ends when something forces the character into the main problem, a long middle of rising complication built around a midpoint that changes what the story is about, and a final act that drives to the climax and settles the cost. Even used loosely, three acts catch the most frequent shape problem, a saggy middle with no turn at its center.

More detailed templates subdivide further. The beat sheet popularized by Blake Snyder under the name Save the Cat lays out around fifteen named beats with rough page positions, including an opening image, a catalyst, a midpoint, an all-is-lost low point, and a finale. The Hero's Journey offers a different set of stages drawn from myth. The Snowflake method, devised by Randy Ingermanson, runs the opposite direction, starting from one sentence and expanding it step by step into a paragraph, then a page, then full scene summaries, so structure grows by repeated zooming rather than by filling a fixed grid.

Two cautions. These templates describe patterns that recur in stories; they do not guarantee one, and a book can hit every beat and still be lifeless if no character drives them. And the page numbers are averages, not laws. Treat a template as a checklist of questions (where is my midpoint, what is my low point, what does the climax cost) rather than a schedule you owe the reader.

The fear behind most outlining questions is rigidity, the worry that planning will kill the surprise that makes drafting worth doing. The fix is to treat the outline as a draft itself, revised as often as the prose. Outline two or three scenes ahead in detail and the rest in signposts, so the near road is clear and the far road stays open to whatever the draft discovers. When a scene teaches you something the plan did not know, change the plan to match the book, not the book to match the plan.

Build the outline so changes are cheap. Cards and short scene documents reorder without rewriting, and a scene list edits in seconds. Heavy prose outlines are the most expensive to change, which is exactly why they tend to go stale and get abandoned. Keep the old version, so you can see what you believed and what the draft taught you. The gap between the two is often where the real story is hiding.

There is also a version you write after drafting, called a reverse outline: you go back through the finished pages and write one line per scene describing what actually happens, ignoring what you intended. Laid out as a list, it exposes problems the forward outline could not, such as a subplot that vanishes for nine chapters or a midpoint that never turns. Reading a whole manuscript this way is hard to do by eye, because the imbalances live across the entire book, not on any one page. DraftProse's Reader works on the full manuscript and reports where pacing stalls or a thread goes quiet, one way to build that reverse outline faster. It reads and reports; it never writes the scenes for you, and the structural decisions stay yours.