Story ideas: how to find one worth writing

Strip any novel you love to one sentence and you will find the same three parts: someone specific, in a situation that demands action, facing a complication that makes the obvious action costly. That triangle — character, situation, complication — is the minimum viable story idea. A setting, a theme, a twist or an image can spark one, but none of them is one yet; they become one the moment you ask who is in trouble here, and why can't they just leave?

This is the fastest filter for your own notebook. "What if memories could be taxed?" is a concept. "A memory-tax auditor discovers her own childhood has been seized as payment" is a premise. The question that converts one into the other is always the same: who does this hurt, and what do they do about it?

Length is a property of the idea, not just the telling. A short story idea turns on a single decision, revelation or reversal — one pressure applied to one person, resolvable in a scene or three. A novel idea has compounding trouble: each attempted fix creates the next problem, and the cast is large enough to disagree about what winning means. Trying to stretch a single-reversal idea across 90,000 words is where most abandoned novels come from.

The practical test: list what has to happen between your opening and your ending. Five events is a short story; twenty-five is a novel; fifteen might be a novella, and our guide to how many words in a novel maps where each length sits commercially. When in doubt, write the short version first — a short story is the cheapest prototype a novel idea will ever get.

Every genre has a native engine, and knowing yours tells you where to dig. Fantasy and science fiction run on consequences: take one impossible change — magic is rationed, the colony's clocks disagree — and ask who profits, who pays, and who enforces it. Horror runs on violation: a safe place or person that stops being safe, and rules that punish the sensible. Mystery runs on concealment: someone is lying about the past, and the past is load-bearing.

Romance runs on incompatible wants held by people the reader needs together — the obstacle must be structural, not a misunderstanding a text message would fix. Contemporary and literary fiction run on the gap between what characters say and what they want, usually inside families, marriages and inheritances. When you are stuck, steal an engine across the aisle: one genre's character dropped into another genre's situation is a reliable idea machine, and it is exactly how the generator above builds its stranger batches.

The escalation test: can you immediately see the second act of trouble — not just the setup but what goes wrong after the first fix? Ideas that pass keep generating scenes; ideas that fail stall at chapter three. The retell test: pitch it to someone in two sentences, then notice which detail you lead with. Whatever you instinctively put first is the live wire; if you lead with the setting every time, you have a world, not yet a story.

And the endurance test, which matters most for novels: you will live with this idea for months. Does it touch something you actually care about — a fear, a grudge, a fascination — or is it merely clever? Clever survives a fortnight. Once an idea passes, give its protagonist a face with a character profile and shape the trouble with an outline before the momentum cools.

Professional writers rarely hunt ideas from a standing start; they harvest a pipeline. Keep one place — notebook, file, the starred shortlist in the generator above — where every premise-shaped idea lands, and date them. When a draft ends, shop the pipeline rather than the void. Ideas also compost: two weak premises from different months often fuse into one strong one, which is a thing no single brainstorming session can do.

And write the ones you fish out. An idea used teaches you what your ideas are missing; an idea hoarded teaches nothing. The notebook is a larder, not a museum.