Turning an idea into a book you can write
An idea is a situation. A novel is a sequence of events, each one caused by the last. Most first novels stall in the gap between those two things, well into a draft, and the author concludes they have run out of talent when what they have run out of is structure.
This is the fiction counterpart to Outlining a nonfiction book so it holds together. Do not borrow that page's shapes and chapter test for a novel; they are built for a book that argues something to a reader, and a novel works on a reader by a different mechanism entirely.
A situation is not yet a story
A woman inherits a house from an aunt she never met. That is a good idea. It is also motionless. You can hold it in your head for two years, describe it well at a party, and still have nothing to write on Monday morning, because nothing in it says what happens next. Ideas feel like books because they carry atmosphere, and atmosphere is the part of a novel that arrives first and does the least work.
What an idea is usually missing
Three things, and it is nearly always the same three.
Somebody who wants something. Not a mood, a wound or a backstory — a want concrete enough to be pursued on a Tuesday and to be failed at. To find out who her aunt was is a topic. To prove the house was not left to her by mistake, before her brother's lawyer files is a want, and it generates scenes by itself.
An obstacle that is not bad luck. Weather, illness and accidents can start a book but cannot sustain one, because the reader learns that anything may happen and stops predicting. The obstacles that work are another person who wants something incompatible, or the protagonist's own character standing where they need to walk.
A reason it starts today. Why this week, rather than a year earlier? If nothing has changed, you are writing the chapters that come before the book begins, and those are the chapters first novels are most often abandoned inside. Something arrives, expires, dies, or is discovered. That is your first scene, and it sits later in the story than you think.
The minimum useful plan
If outlining kills the book for you, plan these four things and nothing else. It takes an afternoon.
The ending, or at least what the book is arguing. Not the last scene — what the book will have claimed about people by the time it stops: that some debts cannot be paid, that the sister was right. Without it every scene is equally plausible, which is the same as having no reason to pick one.
The want and the need, and how they conflict. The want drives the plot; the need is what the book is about. They should pull against each other, so that getting the want costs the need or the reverse. If they are the same thing, your ending is a delivery rather than a change.
Three or four major turns. Moments after which the situation cannot go back. Not chapter summaries — hinges.
The first scene, concretely. Who, where, and what they are trying to do in the next ten minutes.
That is enough to write toward while leaving everything between the hinges undiscovered, which is where the good material comes from.
If you like outlining, two cautions
A scene-by-scene plan can be a way of avoiding the draft. Outlining is writing-shaped, it feels productive, and it produces no pages. The honest question is whether you are outlining because you do not yet know, or because you do and would rather not find out whether you can execute it. Six months of planning documents is a recognizable way of not writing a novel.
Outlines are hypotheses. They are tested by contact with the draft, and they will be wrong, because you do not know what a scene is until you have written it. When scene thirty-one refuses to work, the cause is usually an error upstream in the plan. Authors who defend the outline against the draft end up with a book that is on-plan and inert.
The middle is where books die
A first act more or less writes itself; an ending is exciting to reach. The middle is a hundred pages with no external scaffolding, and it fails in one specific way: it becomes a sequence rather than a story. Things happen, in order, and any two chapters could swap places without damage.
The cure is almost never another event. Adding incidents to a dead middle produces a longer dead middle. The actual fault is that consequences are not landing. The protagonist does things and gets away with them. The opposition does not respond. A revelation in chapter twelve changes nothing about how chapter thirteen is played.
Ask of every middle chapter: what is harder now than it was at the start of it, and what option has been closed off? If nothing has narrowed, the chapter is happening rather than escalating. Tighten the consequences instead of inventing incidents, and the middle usually gets shorter as well as better.
In the Interior Formatter. Previews are free and unlimited, so a finished draft can be set as real pages — chapters detected from the file, a contents built from the actual chapter structure — before you pay for anything. Reading your book as pages rather than as a scrolling document is how a lot of authors first see a middle that does not move. Start your book → · about the Interior Formatter
Two tests worth running
One sentence. Say what the book is about in a sentence containing a character, a want and an obstacle. If it takes a paragraph, the usual reason is that the want has not been decided yet. Tightened, that sentence is also most of your blurb later — see How to write a blurb that sells the book.
And so, not and then. Read your major turns in order and put and so between them. Wherever and then is the honest word, the joint is not load-bearing: the second turn follows the first in time but is not caused by it, and a reader will feel that as the book losing its grip.
Neither test makes a book good. Both are cheap, and both catch the failure that costs a year of drafting.
Then write it and find out
The draft is the experiment, and its answers are better than any you can reason your way to in advance — which is also why the plan should be small enough to throw away. For what the finished thing should weigh in your genre, see How long should a book be?; for the readers who tell you where the middle sagged, Beta readers: what to ask, and when.
Make your book
The tools are free to use: name your book, bring your manuscript or your artwork, and work as long as you like. You only pay when you want the files without the preview watermark.
Next: Outlining a nonfiction book so it holds together · More: Chapter openings and scene breaks · How long should a book be? · The series bible, and what it saves you