How Much Research Does Fiction Need?
Quick answer: How much research does fiction need? Enough to be plausible where it counts: research the load-bearing details, imagine the rest, and show a tenth of it.
Fiction needs enough research to stay plausible at the points where a knowledgeable reader is watching, and no more. Identify the load-bearing details, the ones the plot or the period actually rests on, get those right, and let imagination carry the rest. The working ratio most novelists arrive at is Hemingway's iceberg: know ten times more than you show, but show a tenth of what you know. Research serves the story; the story does not exist to display the research.
Calibrating this matters in both directions. Under-research produces the false note that snaps a reader's trust, a revolver with a safety catch, a Tudor character checking a clock that does not exist yet. Over-research produces something worse: the manuscript that never gets written, because the file of notes kept growing instead.
It Depends on the Contract
Every story signs an implicit accuracy contract with its reader, and the terms vary by genre. Historical fiction and hard science fiction sign strict contracts: readers arrive with expertise and check the plumbing. Hilary Mantel spent years inside Tudor sources before Wolf Hall, and Andy Weir worked real orbital mechanics into The Martian, because their readers would audit exactly those things. A fable or a comic novel signs a loose contract, and nobody audits the botany in a fairy tale.
Within any genre, the strictness also varies by proximity to the plot. If your character merely walks past a foundry, an impression suffices. If the climax turns on how the foundry catches fire, you owe the reader real knowledge. Rank your story's facts by load: what breaks if this is wrong? Research down the list until the load-bearing items are solid, then stop.
Contemporary settings sign contracts too, which surprises writers who assume research belongs to castles and spaceships. Get a working nurse's shift pattern wrong, or a courtroom's procedure, and those readers exit as fast as any Tudor specialist. The present is simply the period you have not noticed you are writing about, and a day of shadowing or one honest interview usually covers it.
How Working Writers Handle It
Patterns recur across desks. Most novelists split research into two phases: a bounded immersion before drafting, enough to move confidently inside the world, and a targeted mop-up after the draft, answering the specific questions the story actually raised. That order matters, because you cannot know which facts you need until scenes demand them.
The other shared habit is refusing to stop mid-draft. When drafting writers hit an unknown, they type a bracket, what does a night nurse carry, and keep going, resolving the brackets in revision. Melville could load Moby-Dick with cetology because he had lived it, but even he shaped the fact-chapters around the story's rhythm. The failure mode has a name in every workshop: the research dump, where the author's homework arrives undigested and the narrative stands still to admire it.
A Practical Research Method
A sequence that keeps research in its place:
The Traps on Either Side
Research-as-procrastination is the respectable trap: it feels like work, produces no pages, and can absorb years. The tell is that your reading has stopped generating scenes. When new material only confirms what you already knew, the immersion is done and the draft is overdue.
The opposite trap is contempt for the reader's knowledge. Getting a famous street, a legal procedure or a horse's gait wrong costs you exactly the readers who care most about your subject, and one broken detail poisons trust in a thousand sound ones. The economical insurance is precision in small things: readers who catch you being right about obscure specifics will extend credit everywhere else. That earned trust, more than any fact, is what research buys.
What poets can take from this
- Rank facts by load: research hard what the plot rests on, lightly what it brushes past.
- Time-box pre-draft research and switch to writing on the deadline, ready or not.
- Bracket unknowns mid-draft; never break momentum to check a fact.
- Hunt texture in primary sources; summaries give you nothing a rival could not also copy.
- Cut any researched detail that does no dramatic work, regardless of what it cost you.
- Get one expert reader to audit the specialist chapters before submission.
Try it yourself
Pick a trade you know nothing about, falconry, tugboats, watch repair, and give yourself ninety minutes of research, no more. Then write a scene where a character in that trade makes a mistake only an insider would notice. Use at most five researched details, and make each one carry story weight.
Questions
Can you write historical fiction without being a historian?
Yes; most historical novelists are self-taught researchers who read deeply for one book at a time. The craft lies in selecting lived-in details rather than displaying complete knowledge, and in getting the era's attitudes right, which readers feel more than dates.
How do you research a novel without getting overwhelmed?
Research in two passes: a time-boxed immersion before drafting, then targeted lookups for questions the draft raises. Keep notes in one searchable file, and let the story, not curiosity, decide what you still need.
Do readers really notice small factual errors in fiction?
The readers who know the subject notice instantly, and they are often the book's natural champions. Most readers forgive small slips if the story holds, but trust erodes cumulatively, so accuracy in checkable details is cheap insurance.
By the Writory editorial team, reviewed by working poets. Updated July 2026.
Enter a Writory poetry contest
Enter the August contest and have your own poem read and scored.